June 12, 1973
Page 19183
NONDEGRADATION
Mr. MUSKIE. Mr. President, yesterday, the U.S. Supreme Court failed to overturn a lower court decision and thus preserved a critical element of the Clean Air Amendments of 1970, the so-called nondegradation provision.
The concept of nondegradation is an essential element to the Nation's clean air effort.
It provides a means to assure that maximum effort will be made to protect air quality from further deterioration.
It requires States to require the best emission control technology available and then take another look to assure that available technology will not result in significant deterioration of air quality.
It provides a means to force development of new, better technology and it provides an interim regulatory mechanism where new source performance standards either do not exist or have not been updated to reflect new control technologies.
Nondegradation imposes a benchmark for State and Federal environmental and planning agencies in making land use decisions, especially siting decisions..
Finally, it establishes firmly that research and development on control technology is to be focused on recycling of pollutants and confined and contained disposal of pollutants, and not on ways of putting pollutants into the ambient environment.
Mr. President, for the past 2 years there has been disagreement between the Subcommittee on Air and Water Pollution and the Environmental Protection Agency regarding the control strategy options available for implementing air quality standards. The Supreme Court action should end that disagreement. Proposals by the Environmental Protection Agency to authorize pollution control strategies based on meteorological and climatological conditions, rather than technological options, should now be shelved.
So-called intermittent control strategies which place reliance on wind, rain, and weather – which require fuel changes or plant shutdown where pollution peaks occur – have always been available for pollution alerts, they do not provide for a basis for regulation; they are not enforceable, and they are no substitute for constant emission controls. Clearly, they are inconsistent with a nondegradation policy because such strategies would permit constant deterioration of air quality.
Tall stacks, another strategy under consideration in the Agency, is also in fundamental opposition to the Clean Air Act and its nondegradation policy. Tall stacks are but manifestations of the out-of-sight-out-of-mind mentality of earlier times. They shift pollution problems to more and more extensive areas of the biosphere.
Mr. President, the Environmental Protection Agency has been considering proposals to permit intermittent control strategies – so-called closed-loop systems – and tall stacks as substitutes for emission controls. Even though this alternative has been mooted by the Supreme Court actions of yesterday, I believe the public should have an opportunity to review the available documents on this strategy.
I ask unanimous consent that a draft document of intermittent control strategies, tall stacks and associated material be printed in the RECORD. Also, I ask unanimous consent that appropriate sections of the Subcommittee on Air and Water Pollution hearings on the viability, enforcability, and legality of intermittent-closed loop-control strategies be printed in the RECORD.
There being no objection, the material was ordered to be printed in the RECORD, as follows:
STATEMENT OF WILLIAM H. RODGERS,. Jr.,
PROFESSOR OF LAW, UNIVERSITY OF WASHINGTON
(Subcommittee on Air and Water Pollution Hearings, Implementation of Clean Air Act – February 16, 1972)
"The closed loop is a sorry strategy for keeping intact smelting technology that poses unacceptable air pollution risks: It is an excuse for avoiding the emission controls the 1970 amendments mandate. Applied to all industries, it would reduce the air pollution regulatory effort to a sham. It is a lawyers' paradise of uncertainties in meteorological prediction, instrument calibration, reading of ambient data and sorting out of SO2 sources, which already has bogged down thoroughly State and local agencies."
* * *
Senator EAGLETON . ... could the closed loop be considered as a control strategy under the 1970 Act as you read it?
Mr. RODGERS. No. I believe that the Act calls for emission limitations, and I submit that the closed loop is not an emission limitation. I might say further on that, because the industry has worked so closely in their technical activity and their political activity with respect to these standards, the word "closed loop" has become almost a catch word in their presentations.
Everyone uses it. Everyone makes the same argument before the different State agencies. As pointed out, basically it is an opportunity to curtail when the weather turns bad. According to their statements, they have been doing that for a century, I think. At least they have been doing that a good part of this century.
Senator EAGLETON. You stated it is not an emission limitation. Could it be considered as a pollution reduction enforcement technique?
Mr. RODGERS. I don't believe so. I think that essentially the problem is that it is unenforceable, and that might be a question put to EPA. I understand that presently in-house there is a draft or at least the agency is considering what might amount to an enforcable closed-loop system as described here. I submit that it is virtually unenforceable.
Senator EAGLETON. How would you apply a compliance schedule to a closed-loop system and how would you monitor it?
Mr. RODGERS. Again it is impossible.
STATEMENT OF BENJAMIN WAKE, ADMINISTRATOR, DIVISION OF ENVIRONMENTAL SCIENCES, MONTANA STATE DEPARTMENT OF HEALTH AND ENVIRONMENTAL SCIENCES, FEBRUARY 16, 1973
The crux of the petition by the Anaconda Co. and the American Smelting & Refining Co., for less stringent ambient air quality standards and for the destruction of the 90 percent emission control standard is, in my opinion, whether the Government, State, and Federal, will be required to use ambient air quality standards as the determining factor in whether or not pollution controls will be applied at all and the effectiveness of such control devices if they are.
The philosophy of the emission standards, on the other hand, revolves round requiring control in keeping with the "most advanced state of the art" which may, in fact, produce ambient air quality better than demanded by the air quality standards.
In my opinion, the latter is the only acceptable posture for the air pollution control program to assume since developing the program around ambient air quality standards procedures acknowledges that control equipment less effective than is currently available or even may have been the standard operating procedure for years, may not be needed simply to roll back to, or not pollute beyond, the ambient standards.
The ambient air quality approach acknowledges that control facilities may not be required at all to prevent the ambient air quality standards from being exceeded. Adopting the ambient air quality standard philosophy as the primary determination of whether or not control devices will be installed is to permit degradation of air quality that is better than the ambient air quality standard and to require rollback of air dirtier than the standards only to that standard and no more.
The latter, in fact leaves no margin for release of unavoidable emissions from other emission- producing enterprises that may come into the area. Once the ambient air quality standard is reached, there is no way to get – but dirtier.
STATEMENT OF FRANK MILLIKEN, PRESIDENT,
KENNECOTT COPPER CORP., FEBRUARY 24, 1972
Senator EAGLETON. Are you going to apply a closed-loop system?
Mr. MILLIKEN. Yes, we will have a closed-loop system for surveillance of our operation so anybody can see what sulfur dioxide concentrations are at numerous places around our properties.
Senator EAGLETON. Who will monitor those?
Mr. MILLIKEN. We will monitor those, but they can be monitored by the States if they wish.
Senator EAGLETON. In a State like Montana, which doesn't have the most expansive budget in the world, is there expected to be a State agent to monitor?
Mr. MILLIKEN. Not if they will take the word of the company's operations, although this stuff will be on computer printout. Of course, you could juggle those if you wanted to, and someone could make that accusation, but we don't expect that to happen.
Senator EAGLETON. You will purchase the monitoring equipment and your employees will do monitoring?
Mr. MILLIKEN. That is right. That is what we propose. If someone wants something different, we will have to listen.
STATEMENT OF CHARLES BARBER, CHAIRMAN,
AMERICAN SMELTING & REFINING CO., FEBRUARY 24, 1972
The curtailment of operations that is currently required to implement the intermittent control systems at our copper plants has been costly to us and the mines that ship to us. Copper production at our El Paso smelter was reduced 29 per cent by air pollution curtailment during 1971. We expect, however that by 1974 the need to curtail production, even to meet the federal secondary standards, will be minimal. I say this because, in order to achieve the goal of timely and full compliance with federal ambient standards, Asarco is investing $50 million in sulfur removal facilities at our three copper smelters. These installations will recover more than 50 percent of the process sulfur at each plant.
ENVIRONMENTAL PROTECTION AGENCY,
OFFICE OF THE GENERAL COUNSEL,
Washington, D.C.,
April 30, 1973.
Reply to: Michael A. James, Attorney, Air Quality and Radiation Division.
Subject: Implementation of Section 110 of the Clean Air Act.
To: Joe Padgett, Director, Strategies and Air Standards Division, Office of Air Quality
Planning and Standards, OAWP.
MEMORANDUM OF LAW
Facts
Your memorandum of February 27, 1973 to Robert Baum raises several questions involving subjects discussed at the Regional Administrators' meeting on power plants. All of the questions are concerned with EPA's overseeing of State implementation plans.
Question No. 1
If a State has an approved emission regulation which is more stringent than necessary to attain the national standards but refuses to enforce its emission regulation by obtaining compliance schedules from regulated sources, may EPA reject the State emission regulation and promulgate a less restrictive measure that provides for the attainment of ambient air quality standards?
Answer No. 1
Where EPA has approved a State emission regulation as part of an applicable plan and the State does not enforce the regulation, EPA's responsibility under the Clean Air Act is to enforce the approved emission limitation and, in so doing, the Agency must provide for compliance with the approved emission limitation.
Discussion No. 1
It is helpful to begin with a general discussion of EPA's authority and responsibility under § 110 and 113 of the Act, since most of the questions raise basic problems of interpretation of those sections. It is important to recognize that we are discussing two separate functions, viz approval/promulgation and enforcement.
EPA's authority to promulgate implementation plan regulations stems from the disapproval of regulations submitted by the State, or by the failure of the State to submit necessary regulations.
If State regulations are approved by EPA, the Agency has no authority to promulgate different regulations. Under the law, EPA must approve regulations which are more stringent than those needed to meet the national standards. Once these regulations are approved, there is no authority to promulgate less stringent regulations. This is true even if a State fails to enforce these regulations.
With regard to the second function raised by the questions; i.e. enforcement, EPA is given clear authority to enforce approved implementation plans or plans promulgated by the Administrator. As we have previously pointed out, under § 110(d), for purposes of the Clean Air Act "... an applicable implementation plan is the implementation plan, or most recent revision thereof which has been approved under subsection (a) or promulgated under subsection (c) and which implements a national primary or secondary ambient air quality standard in a State." The words "applicable implementation plan" are in this case, words of art. Section 113 authorizes Federal enforcement of an "applicable implementation plan." Accordingly, it is clear that it is only approved or promulgated plans which EPA may enforce.
As you know, the submission by a State with regard to regulations and compliance schedules is really two separate submissions. On one hand, EPA evaluates the emission limitations to make certain that they are sufficient to achieve the national standards. If the degree of reduction is sufficient, that emission standard is approved. Many State plans contain provisions by which they are required to procure a compliance schedule subsequent to the adoption and submission of the emission standard. Failure to obtain the compliance schedules in no way affects the validity of the approved emission regulation. Accordingly, EPA does not have authority to promulgate a different emission regulation. What is left to EPA is the authority to procure compliance schedules which meet the applicable implementation plan, in this case, the emission limitations submitted by the State and approved by EPA.
Question No. 2
When imposing Federal compliance schedules or approving State compliance schedules for sources subject to approved State emission regulations which are more stringent than necessary to attain the national standards, must EPA require compliance with the approved regulation or may it impose or approve instead whatever less stringent requirements are necessary to achieve the national standards?
Answer No. 2
Unless the State revises its approved regulation and obtains EPA approval of that revision, both the State and EPA are bound by the approved regulation when obtaining or approving compliance schedules.
Discussion No. 2
The premise of your second question is that the State has submitted emission limitations which are more stringent than necessary to achieve the national ambient air quality standards. The issue is whether if a State submits a compliance schedule or we have to procure one, can we accept or procure one which will achieve the standards or must we accept or procure one which meets the State emission regulations. This situation is similar to the first one discussed above. The applicable plan contains an emission limitation which is the only guide for the preparation and approval of compliance schedules. Quite aside from the requirements of § 110, a different answer would put EPA in the position of approving or trying to secure a compliance schedule to meet an emission limitation which does not exist, except in EPA files. More specifically, even if it were possible to try to adopt or procure compliance schedules to meet some number less stringent than that approved in the plan, exactly what that number would be in each case would be subject to question and litigation. We should point out that if the State has in fact adopted emission limitations which are more stringent than necessary to meet the national standards, they can submit a plan revision with more lenient requirements if they still conform with the requirements of the Act.
Question No. 3
Is a change in control strategy by a State (e.g. from a firm emission limitation to a system of intermittent control, tall stacks, and/or some other measures) to be considered a plan revision?
Answer No. 3
Yes. This action would constitute a substantive modification of the regulatory scheme which carries out the control strategy to provide for attainment and maintenance of the national standards.
Discussion No. 3
The change in question would involve the regulatory requirements applicable to a source or class of sources. Emission limitation requirements are the most critical parts of any plan and are specifically required to be included in the plan by § 110 (a) (2) (B) of the Act. It is axiomatic that a substantive modification of such requirements must be considered a plan revision.
Question No. 4
May States revise an approved plan requirement because of the difficulty or impossibility of sources meeting that requirement? Where a State makes such a determination, may it now apply for an extension of the statutory attainment date for the national standards?
Discussion No. 4
Where the State, in negotiating compliance schedules with individual sources, determines that compliance with the approved emission regulation by a source or sources will be difficult or impossible by the prescribed compliance date, it may revise its plan with respect to that source or sources. A source may be granted a variance from the initially-applicable compliance date if compliance is required to be as expeditious as practicable (40 CPR 51.15(b)) and the compliance date does not extend past the prescribed attainment date for the national standards. Any extension of compliance past that date would require a postponement under § 110(f) of the Act (40 CFR 51.32(f)).
Alternatively, the State may reassess the control strategy and choose to revise its emission regulations to reflect the non-availability of technology or other control measures (e.g. low sulfur fuels), if the revised regulations will still provide for attainment of the national standard within the prescribed attainment date. The State may also set back the attainment date for a national standard if the new date is no later than mid-1975 and the plan demonstrates that the new date represents attaining the national standard as expeditiously as practicable.
Question No. 5
May EPA approve implementation plan provisions which utilize stack height requirements for emission dispersing in lieu of measures requiring limitation of emissions?
Answer No. 5
As noted in your memorandum, this question is now being considered by the Court in the National Resources Defense Council suit challenging EPA's approval of the Georgia plan, and we feel it is appropriate for us to defer any action on the question until the Court makes a decision.
Discussion No. 5
As you may be aware, a briefing package on the stack height limitation issue is being prepared for the Administrator's consideration.
Question No. 6
Does the Act allow a State revise a plan by acquiring emission regulations adequate to attain the national standards but less stringent than those approved by EPA or to require emission regulations resulting from a reclassification of a region from Priority I to Priority III?
Answer No. 6
Yes, provided the State demonstrates to the Administrator's satisfaction that the less stringent regulations provide for the attainment of the relevant national standards as expeditiously as practicable, but no later than mid-1975. In the case of regional reclassification, the Administrator could approve the recission based on a determination that the controls are not necessary since the national standard (NO2) is being attained. Where the standard is being attained only marginally, however, recission of all NOx controls may threaten maintenance of the standard and necessitate the Administrator's disapproval of all or part of the recission.
Discussion No. 6
In our view, § 110 did not require States in the preparation of their plans to make faultless judgments with respect to the practicability of controlling sources and attaining the national standards. Reassessments and consequent revisions to plans are approvable by the Administrator so long as the revised plan demonstrates attainment of the national standards as expeditiously as practicable (but no later than mid-1975). As noted in No. 4 above, in the case of individual source compliance schedules (including variances), the source must be required to comply as expeditiously as practicable (40 CFR 51.15(b)). The unavailability of low sulfur fuels is an appropriate factor for consideration in determining the practicability of control, both as applied to individual sources (in compliance schedule development) and to attainment dates.
It should be noted that the Agency is currently engaged in litigation with the Natural Resources Defense Council over the question of relaxation of plan requirements, through either granting of variances or other regulatory revisions. NRDC argues that the only permissible means of postponing plan requirements is pursuant to § 110(f) of the Act, the provision for one-year postponements upon specific findings by the Administrator on the record of a formal hearing.
PROPOSED FEDERAL REGISTER NOTICE RECOGNIZING THAT TALL STACKS AND VARIABLE (INTERMITTENT) CONTROL MAY BE USED FOR SOME SOURCES TO PROTECT AGAINST VIOLATIONS OF SO2 NAAQS – ACTION MEMORANDUM
SYNOPSIS
It is proposed that EPA provide for the use of dispersion enhancement techniques (tall stacks and variable emission or intermittent control systems) in State implementation plans. The attached draft of a Federal Register notice sets forth the conditions under which such techniques may be applied. Their use would be limited to large, existing, remote facilities which cause violations of the SO2 NAAQS.
DISCUSSION
The air quality standards represent very restrictive targets that provide for the protection of public health and public welfare. In order to meet them in a timely fashion and without unreasonable social disruption, it is necessary to utilize all the techniques available to the air pollution control profession and to constantly seek new techniques that will hasten attainment, and lower the social impact of achieving clean air, yet not sacrifice the integrity nor credibility of the Clean Air Act.
Among the technological approaches EPA has been examining for several years are techniques which take advantage of the capability of the atmosphere to disperse and dilute pollutants. There are two approaches – increasing the effective height that the emissions take place, i.e., tall stacks, and managing the rate of emission according to the continually changing dispersive capability of the atmosphere (often called intermittent control or variable emission control). It is now concluded that for a limited number of cases and under carefully controlled conditions, these dispersion enhancement approaches can be added to the techniques available to control air pollution and to meet ground-level ambient air quality standards.
In arriving at this conclusion, it is recognized that there is value in reducing the pollution load on the atmosphere by removing emissions rather than relying wholly on dilution techniques to meet air quality standards; therefore, these techniques should be considered only where adequate emission control technology is not readily available or reasonable to apply. Generally, effective and relatively inexpensive techniques are available for the control of particulates and CO, and sources emitting these pollutants are not likely candidates for dispersion enhancement. It should be noted that once effluents leave any source, the natural dispersion and removal processes of the atmosphere dilute the concentrations. All strategies designed to attain national ambient air quality standards take advantage of these processes in some way.
It is also recognized that a dispersion enhancement system must conform to the same tests of reliability, enforceability, and source responsibility that are applied to more conventional air pollution control strategies. Therefore, at this time, techniques to attain standards by enhancing dispersion are being considered only for isolated sources whose impact on air quality is unambiguous and when terrain, meteorology and the source location makes relatively simple the design and enforceability of effective variable emission control systems. This emerging technology cannot presently accommodate systems involving hydrocarbons, oxidants or nitrogen dioxide. Because of the atmospheric reactions involved with these pollutants, the knowledge required to relate emissions from a specific source to air quality throughout the area is simply not available. Similarly, the application of these techniques for any pollutant simultaneously to many sources, especially in or around metropolitan areas, is tenuous and cannot be reliably enforced at this time.
Because variable control systems have been discouraged in the past, data on their reliability is sparse. Recently, however, TVA and ASARCO presented data from three widely separated geographical areas (Kentucky, El Paso and Tacoma) indicating that these approaches can significantly reduce violations of the short-term SO2 standards. The data are "company" data, but the Tacoma information is generally supported by independent data collected by the Puget Sound APCA. On the basis of these data, and an assessment of the reliability of diffusion modeling and emission reduction techniques, it is now concluded that these techniques may be used by some sources to protect against violations of SO2 standards as effectively as flue-gas control systems.
The use of techniques to take advantage of the dispersive capability of the atmosphere is subject to inherent uncertainties due to its great and often rapid variability in space and time. Therefore, it seems prudent at this time to discourage their use for attaining primary air quality standards – those that are designed to protect public health. There are three problems that attend the implementation of such a policy. The first problem is that limestone scrubbers are not
particularly applicable. If a source must reduce emissions by 40% to meet primary standards, and 80% to meet secondary standards, and no means other than a limestone scrubber are available, then the option to use dispersion enhancement for secondary standards is effectively foreclosed.
The second problem is that emission reduction control methods (cleaner fuel, scrubber, acid plant) may be unavailable, insufficient, or infeasible for meeting the primary 24-hr standard in some instances. The choice would then be between forcing plant operation curtailment or shutdown, and granting a variance. Dispersion enhancement would be preferable to either of these options.
The third problem is that considerable emphasis has been placed on insuring that dispersion enhancement systems will be designed, operated end enforced so that all air quality standards will be reliably achieved. Since secondary standards will be attained, the less stringent primary standards will simultaneously and reliably be attained. Therefore, the basis for rejecting dispersion enhancement for meeting primary standards no longer exists.
The decision as to whether dispersion enhancement may be used for a particular source is built into the proposed regulations up to the point where no more than 150-200 large, isolated, existing SO2 sources can qualify for consideration. Although an element of judgment is inescapable when weighting the various inputs to the acceptability decision, the information required by Appendix Q should allow clear decisions in the public interest in most cases. It is inevitable that the limited acceptance of dispersion enhancement will lead to a law suit. One issue in such a suit is the interpretation of the Clean Air Act.
The Clean Air Act states (Sec. 110(e) (2) (B)) that each State Implementation Plan must include "emission limitations, schedules and timetables for compliance with such limitations, and such other measures as may be necessary to insure attainment and maintenance of such primary and secondary standard, including, but not limited to, land use and transportation controls." This requirement, and the words "and such other measures" in particular, may be interpreted in three ways : (1) Emission limitation is the only allowable means of attaining standards; "other measures" are alternative means and procedures for effecting emission limitation; dispersion enhancement is not acceptable, (2) "other measures" refers to measures other then emission limitation (e.g., dispersion enhancement). Such measures are allowable if they are necessary (i.e., if emission limitation sufficient to attain standards is unavailable or impractical), (3) any combination of measures which attain and maintain national standards is acceptable provided emission reduction is included.
The strong emphasis on emission reduction throughout the Clean Air Act (see Sections 111(a)(1), 111(d)(1) and 112(b)(1)(B)) and the benefits of emission reduction over dispersion enhancement lead OAWP to conclude that the third interpretation is environmentally unsound and inconsistent with the intent of Congress. On the other hand, the fact that dispersion enhancement can reduce ground-level concentrations at moderate cost and acceptable reliability while some emission reduction methods for sulfur oxides are expensive, limited in availability, based on emergent technology, and of only moderate efficiency and reliability argues strongly for the inclusion of dispersion enhancement in the array of acceptable control techniques.
Consequently, Section 110 (a) (2) (B) is interpreted to mean that dispersion enhancement is an "other measure" which may be used when "necessary." This legal interpretation underlies this proposed change in the regulations.
Issues which are expected to arise include:
a. Is it legally and technically possible to limit use of dispersion enhancement techniques to attainment of SO2 standards by large isolated power plants and smelters? Technically, it is appropriate to confine the use of these techniques to large isolated sources. Responsibility for the violations is easily shown; enforcement is simplified. Non-urban areas allow flexibility in acquiring the large amounts of data necessary to develop and demonstrate the reliability of the procedures. The impact of threats to the standard is more readily assessed. The legal arguments for limiting the use of these procedures are not known. A suit should be anticipated.
b. Whet other types of industries might desire to apply these procedures? Sulfuric acid plants and zinc and lead smelters are potential candidates for use of these procedures to attain SO2 standards. If they meet the isolation test and they are located so that the controls needed to meet the standards are infeasible or unavailable, they warrant consideration.
c. Should oil-fired power plants be considered? No. The control technology, de-sulfurized fuel, is prevalent and highly reliable. Further, few, if any, oil-fired plants are isolated. In this regard, EPA will continue to discuss with the concerned firms the technical, legal and enforceability aspects of the Pioneer Valley, Long Island Lighting Company end General Electric (Lynn, Mess.) proposals.
d. What increase in total emissions of SO2 should be expected? The change in total emissions should be limited to the increase caused by additional demand for energy placed on existing generating facilities. New sources will come under new source performance standards. Emissions will be managed so that adverse effects on ground-level air quality will be minimized.
RECOMMENDATION
That you approve the enclosed revisions and additions to 40 CFR Part 51.
Appendix Q, description of an Intermittent Control System and Criteria for a Regulation.
This appendix describes procedures to supplement the attainment and maintenance of National Ambient Air Quality Standards for sulfur dioxide by taking advantage of the dispersive capability of the atmosphere. The air quality standards represent very illusive targets that provide for the protection of public health and welfare. In order to meet them in a timely fashion and without unreasonable social disruption, it is necessary to use all the techniques available to the air pollution control profession and to constantly seek new techniques and to reevaluate and upgrade old techniques that will hasten attainment and lower the social impact of enhancing the air environment. Among the technological approaches being examined are systems that more fully use the dispersive capability of the air. It is now concluded that for a limited number of cases and under carefully controlled conditions, procedures which enhance the dispersion of effluents from large isolated existing sources of sulfur dioxide can be added to the techniques available to meet ground-level ambient air quality standards. In making this determination, it is recognized that there is value in reducing the pollution load on the atmosphere by removing emissions rather then relying wholly on enhancing dispersion to meet the air quality standards. It is recognized that continuing and increasing demands for energy place increasing stress on the environment, including the quality of the air, and that even under the best conditions for dispersion, the dispersive capability of the atmosphere may be overwhelmed. Therefore, use of techniques to enhance dispersion can be considered only where adequate emission control technology is not readily available or reasonable to apply.
The statements presented herein are not intended, and should not be construed, to require or encourage State agencies to authorize procedures to enhance the dispersion of sulfur dioxide as a means to attain and maintain air quality standards without considering (1) the frequency and severity of threats to the air quality standards in the vicinity of large, remote sources of sulfur dioxide, (2) the availability and socio-economic cost of emission reduction control technology for the attainment of air quality standards around such sources, (3) the availability of low sulfur fuel, (4) the reliability and enforcement problems associated with dispersion enhancement techniques, and (5) the environmental effects of emissions even though such emissions are sufficiently diluted at ground level to attain air quality standards.
Failure of a State agency to adopt a technique for enhancing dispersion of emissions to attain and maintain National Ambient Air Quality Standards within the time prescribed by the Clean Air Act will not be grounds for rejecting a State implementation plan if the plan provides for attainment and maintenance of the standards. Nor will adoption of such techniques be grounds for the approval of the implementation plan if the national standards are not attained and maintained. In preparing plans which use dispersion enhancement techniques, State agencies should be assured that the plans deal with the particular and unique problems of their own State and that the techniques they approve deal with the problems in a reliable and enforceable manner.
1.0 DEFINITIONS
"Intermittent Control Systems" are designed to meet air quality standards by varying the emission rate with meteorological conditions in order to take advantage of the continually changing dispersive capacity of the atmosphere.
"Effective stack height" means the sum of the physical height of the stack above grade and the height the effluent plume rises above the height of this stack top. Under most circumstances, an increase in the effective stack height results in a decrease in the maximum ground-level concentration of the emitted pollutant and an increase in the distance from the source that the maximum concentration is experienced.
2.0 GENERAL CONDITIONS FOR ACCEPTABILITY OF TECHNIQUES TO ENHANCE DISPERSION OF POLLUTANTS
The following general conditions must be satisfied before techniques to enhance dispersion of pollutants may be applied to attain and maintain National Ambient Air Quality Standards.
2.1 Emission enhancement techniques may be applied to sulfur dioxide emissions only.
2.2 The existing source of sulfur dioxide emissions must be remote from other sources (e.g., located in an area where the contribution of other sources does not cause contamination of more than 10% of the annual standard.)
2.3 Emission control technology for the source's sulfur dioxide emissions is unavailable, infeasible or insufficient to attain and maintain the National Ambient Air Quality Standards or would result in unreasonable social disruption.
2.4 The technique to enhance dispersion of sulfur dioxide will enable the National Ambient Air Quality Standards to be met in a timely fashion.
2.5 The technique to enhance dispersion will include intermittent control of the emissions and adequate effective stack height. Increasing effective stack height without the application of intermittent control procedures is not an acceptable technique.
2.6 The technique to enhance dispersion must be reliable and enforceable. It must conform to the same tests of reliability, enforceability and source responsibility as are applied to techniques to attain National Ambient Air Quality Standards by the constant, continuous and permanent control of emissions.
3.0 ELEMENTS OF AN INTERMITTENT CONTROL SYSTEM
3.1 Figure 3.1 presents a block diagram of the elements of an intermittent control system and the relationships among them.
3.2 The function of each element follows:
(a) Meteorological Inputs. Observations and predictions of the values of meteorological parameters required by the operational model to determine the degree of control needed to avoid threats to the air quality standard.
(b) Operation model. An intellectual construct which relates meteorological inputs, emission rates, source data and terrain and location factors to current and future ambient air quality in the vicinity of the source.
(c) Schedule emission rate. The emission rate which would result under the currently scheduled processes and levels of operation.
(d) Control decision. Decisions, based on either the model prediction or real-time air quality, whether or not to continue with scheduled processes and their attendant emissions, and if not, how much to curtail the emission rate.
(e) Controlled emissions. The emission rate resulting from the control decision.
(f) Actual meteorological conditions. The measures of wind speed, wind direction, stability, mixing height and other weather factors at the time of emission release.
(g) Air quality. Actual ground-level pollutant concentrations and their spatial distribution.
(h) Air quality monitors. An array of SO2 sampling stations located at the points where maximum ground-level concentrations are most probable to occur, at representative points which are readily accessible to the public, and, in sufficient numbers to allow calibration of the diffusion model so that it may accurately interpolate air quality between samplers. A portion of the monitors may be mobile or portable.
(i) Threshold values. Values of SO2 concentration somewhat below air quality standards and/or rates of change of SO2 concentrations which serve as indicators of potential violations of the standard. They are selected so that a control decision for emission reduction can be made in sufficient time to prevent air quality standards from being violated.
(j) Data storage. Time phased records of meteorological conditions, emission rate, model prediction, measured air quality and control decisions available for control agency review and model upgrading.
(k) Upgrade model. A periodic evaluation of the model's prediction accuracy and, if possible, a revision of the form or parameters of the model in order to improve that accuracy.
3.3 The intermittent control system described in Figure 3.1 will be seen to consist of three basic operations: control based on air quality prediction, control based on air quality measurement, and periodic model upgrading. Each of these operations is considered necessary to a reliable system, for each performs a valuable function. The operating model is used to predict ground-level pollutant concentration sometime in advance of its potential occurrence, and to interpolate between monitors. The monitored data and threshold values are used to supplement and, if necessary, override decisions based on the model output thus compensating for the less than perfect accuracy of the model. The model upgrade operation is used to convert the tentative initial model into an accurate prediction mechanism tailored to the specific plant and site.
4.0 CRITERIA FOR AN ACCEPTABLE REGULATION AUTHORIZING USE OF TECHNIQUES TO ATTAIN NATIONAL AMBIENT AIR QUALITY STANDARDS BY INTERMITTENT CONTROL SYSTEMS
4.1 This section presents criteria for an acceptable regulation concerning intermittent control systems. The purpose of such a regulation is to ensure that the proposed intermittent control system will enable air quality standards to be attained and maintained, that the system will be reliable and enforceable, and that the necessary elements of the system will be clearly and legally identified.
4.2 An acceptable regulation should:
(a) Authorize approval of each ICS only after reasonable notice and public hearing.
(b) Define air quality violations as:
(1) A single ambient concentration that exceeds the standard at any air quality monitor.
(2) Repeated or consecutive excesses at the same monitor or non-simultaneous excesses at different monitors are multiple violations.
(3) Non-compliance with stated and agreed upon emission curtailment conditions and procedures.
(c) Permit a source to submit a plan for an ICS only after the source justifies the need for the system. The justification should discuss:
(1) Type and location of facility.
(2) Demographic aspects of the location.
(3) Anticipated growth: population, industrial, urbanization, etc.
(4) Frequency and severity of air quality standard violations.
(5) Availability and reliability of constant control systems.
(6) Economic aspects of constant control methods.
(7) Life expectancy of facility.
(8) Plan for development and demonstration of an ICS.
(9) Other factors pertinent to the facility.
(d) Apply only to those sources which are reasonably remote from other sources of the same pollutant (e.g., in areas where the source will assume full responsibility for observed SO2 concentrations).
(e) Apply only to those sources which can curtail their emissions at a rate compatible with the advance warning time (of adverse atmospheric dispersion conditions) afforded by the ICS.
(f) If a permit is granted, require periodic re-justification (e.g., 3-5 year intervals) to insure that changes in economic, control, demographic, etc., factors do not warrant change in authorization for the ICS.
(g) Authorize a fee for the permit (funds from which will be used by the control agency for the additional surveillance and enforcement functions created by the intermittent control system).
(h) Require the source to establish, maintain and continuously operate monitors for sensing the rate of emission of the pollutant, air quality and meteorological parameters.
(i) Grant the agency continuous access to all data generated by the source's network of sensors and authority to inspect, test and calibrate all sensors, recorders and other equipment of the monitoring network.
(j) Require the source to notify the control agency when emission curtailment is initiated and when air quality standards are exceeded.
(k) Authorize the control agency to initiate emission curtailment as it seems appropriate; i.e., allow the agency to override the source's operation of the ICS.
(1) Require the source to submit a plan and schedule for implementing an ICS. The plan shall have two parts:
(1) A comprehensive report of a thorough background study which demonstrates the capability to operate an ICS. The report shall describe a study during a period of at least 120 days when air quality standards are frequently or likely to be exceeded which:
(i) Describes the emission monitoring system and the air monitoring network.
(ii) Describes the meteorological sensing network.
(iii) Identifies the frequency, characteristics, times of occurrence, and durations of meteorological conditions associated with high ground-level concentrations.
(iv) Describes the methodology (e.g., dispersion modeling and measured air quality data) by which the source determines the degree of control needed under each meteorological situation.
(v) Describes tests and results of tests to determine optimum procedures and times required to reduce emissions.
(vi) Estimates the frequency that ICS is required to be implemented to attain air quality standards.
(vii) Describes the basis for the estimate.
(viii) Includes data and results of objective reliability tests. "Reliability," as the term is applied here, refers to the ability of the ICS to protect against violations of air quality standards.
(2) An operational manual which:
(i) Specifies and substantiates the number, type, and location of ambient air quality monitors, in-stack monitors, and meteorological instruments needed.
(ii) Identifies the meteorological situations before and/or during which emissions must be reduced to avoid exceeding short-term air quality standards.
(iii) Describes techniques, methods and criteria used to anticipate the onset of meteorological situations associated with the excessive ground-level concentrations.
(iv) Describes the methodology by which the source determines the degree of control needed for each situation.
(v) Identifies specific actions that will be taken to curtail emissions when critical meteorological conditions exist or are predicted to exist and/or when specified air quality levels occur.
(vi) Identifies the company personnel responsible for initiating and supervising such actions.
(vii) Demonstrates that the curtailment program will result in maintenance of short-term and long-term air quality standards.
(viii) Describes the manner by which monitoring data are transmitted to the control agency (in a manner acceptable to the agency).
(ix) Describes a program whereby the source systematically evaluates and improves the reliability of the ICS.
(x) Identifies a responsible and knowledgeable person (and alternate) who can apprise the control agency as to the status of the ICS at any time.
(m) Require the source to submit monthly reports on the ICS, including an analysis of how the system affected air quality and how response to adverse dispersion conditions will be improved.
(n) Require annual review of the ICS by the control agency, and authorize the agency either to impose a fine on the source or to deny its continued use of the ICS if:
(1) The source has not complied with all provisions designed to protect long-term standards.
(2) The source has not developed a control program that is effective in enabling short-term standards to be met.
(3) The source has not demonstrated good faith in operating an effective control program by failing to:
(i) Utilize trained competent personnel.
(ii) Maintain and calibrate the monitoring equipment properly.
(iii) Refine and continuously validate and upgrade the response of the ICS to adverse dispersion conditions.
(iv) Attain annual and short-term standards in the vicinity of the source.
ENVIRONMENTAL PROTECTION AGENCY
[40 CFR Part 51]
REQUIREMENTS FOR PREPARATION, ADOPTION, AND SUBMITTAL OF IMPLEMENTATION PLANS
Notice of proposed rule making
On August 14, 1971 (36 F.R. 15486), the Administrator promulgated as 40 CFR Part 420, regulations for the preparation, adoption, and submittal of State implementation plans under Section 110 of the Clean Air Act, as amended. These regulations were republished November 25, 1971 (36 F.R. 22398), as 40 CFR Part 51. The amendments proposed herein would revise 40 CFR Part 51, Subpart A, § 50.1 and Subpart B, § 50.12, § 50.13. The amendments proposed herein would also revise 40 CFR Part 51 by adding a new Appendix Q and providing additions to existing Appendix B.
The proposed amendments to 40 CFR Part 51 provide for the use of dispersion enhancement techniques in State implementation plans and set forth the conditions under which such techniques may be approvable.
Dispersion enhancement means the release of pollutants into the ambient air such that those pollutants are distributed throughout a larger volume of air and over a larger land area thus reducing the peak and average pollutant concentration at ground level. There are two basic techniques which produce this effect: (1) temporal variation of emission rate based on the dispersive capacity of the atmosphere as indicated by certain predicted and observed meteorological conditions (e.g., wind speed, stability, mixing height), and (2) increasing the height of the effluent plume through increased stack height or increased effluent temperature and/or exit velocity. These two techniques are complementary and should be employed together if dispersion enhancement is used to achieve air quality standards at ground level.
In the past, emphasis has been placed on attaining standards by reducing emissions. Tall stacks and techniques to vary the emission rate based on weather conditions (variable control systems) have not been encouraged. These approaches deliberately take advantage of the capability of the atmosphere to disperse and dilute pollutant concentrations. Because the dispersive capability of the atmosphere varies over several orders of magnitude with time and location, the reliability of these techniques compared to reducing emissions continuously has been questioned. The enforcement of regulations that would authorize the use of variable control techniques appears difficult, complicated, less certain and more costly than enforcement of regulations requiring permanent emission reduction. In addition, there are clear benefits to society in limiting emissions independent of the achievement of NAAQS. The past position with respect to variable (intermittent) control systems is conveyed in 37 F.R. 10845, May 31, 1972, and specifically in 37 F.R. 15095, July 27, 1972, which states:
"At this time, it (variable control) is not considered an acceptable substitute for permanent control systems for attaining and maintaining national standards. Experience with systems employing intermittent process curtailment indicates that although air quality is improved, violations of ambient air quality standards still occur. Additional experience with these systems may, however, in specific cases improve this reliability.
"(7) All sulfur dioxide emissions are required to be properly captured and vented through a stack. Although this may result in some improvement in air quality, the precise degree of improvement cannot be defined at this time; accordingly, it could not be taken into consideration in determining the total degree of emission control required to attain and maintain national standards."
The acceptability of dispersion enhancement techniques is being reevaluated for three reasons: (1) Recent data for both coal-fired power plants and copper smelters indicate that variable emission control systems can reduce ground-level concentrations to levels below air quality standards with a reliability in some circumstances equivalent to stack gas cleaning devices when such systems are operated in conjunction with adequate stacks, (2) Section 110 (a) (2) (B) of the Clean Air Act specifies that other features besides emission reduction may be used to attain air quality standards if such other measures are necessary, (3) the availability cost and threats to other aspects of the environment associated with emission reduction methods for SO2 are such that other measures (i.e., dispersion enhancement) may be necessary for the timely and cost- effective attainment of air quality standards.
The proposed revisions and amendments to 40 CFR Part 51 restrict the use of dispersion enhancement techniques to isolated sources of SO2. These restrictions are based on the strong preference for emission reduction in the Clean Air Act [see Sec. 110(a) (2) (B), 111(d)(1) and 112(b)(1)(B)], and on the necessity of relating source emissions to the resultant pollution concentrations in order to reliably operate dispersion enhancement systems. Cost-effective emission reduction control technology is considered by the Administrator to be available for stationary sources of carbon monoxide and particulate matter, so dispersion enhancement is unnecessary for control of those pollutants. Source accountability for ground-level concentration of nitrogen dioxide and ozone cannot be estimated with sufficient accuracy for dispersion enhancement techniques to be a reliable means of control for stationary sources of those pollutants. Therefore, dispersion enhancement will be acceptable only for sources of sulfur oxides.
In order for dispersion enhancement to be reliably operated and adequately enforced, the ground- level concentration of sulfur dioxide must be related to the source emission rate. In urban or other areas where many sources may contribute to the observed ground-level concentration, the emission concentration relationship of any one source is difficult to estimate with sufficient accuracy to allow reliable emission control or supportable enforcement action in the event that air quality standards are violated. Therefore, dispersion enhancement will be acceptable only for isolated sources which will accept full responsibility for ground-level concentrations of sulfur oxides in their vicinity.
Two additional conditions must be met for dispersion enhancement to be acceptable for non-urban sources of SO2 emissions: emission enhancement techniques must be necessary to attain air quality standards, and the proposed emission enhancement must be technically capable of achieving air quality standards with a reliability consistent with emission reduction methods.
Demonstration of the necessity for emission enhancement will be made if (1) All available and practical emission reduction means have been employed, (2) air quality standards are threatened by the residual emissions, and (3) further emission reduction means are unavailable, infeasible, would result in serious socio-economic disruption, or are impractical for other reasons. The fact that emission reduction may be more costly than dispersion enhancement is not necessarily a sufficient demonstration of the necessity for dispersion enhancement, although cost considerations are clearly germane to this demonstration. For example, oil-fired power plant emissions of sulfur dioxide may be controlled by the use of desulfurized oil. Dispersion enhancement is not necessary for oil-fired power plants even though it may be less expensive than the use of desulfurized oil. On the other hand, the use of limestone scrubbers on coal-fired power plants to achieve short-term air quality standards which are violated less than one percent of the time may be over ten times as costly as variable emission control. It may be no more reliable and may seriously degrade the environment by producing large amounts of liquid and solid waste. Such a case would clearly qualify for consideration of dispersion enhancement.
Certain nonferrous smelters may also be candidates for dispersion enhancement control techniques. Acid plants are cost-effective control methods for removal of the majority of sulfur from the emissions of such sources, but this emission reduction may not be sufficient for standard attainment. Further control of ground-level sulfur dioxide concentration using dispersion enhancement may be acceptable in such instances.
The determination of when dispersion enhancement is "necessary" cannot be made with precise objectivity. The problem is to balance the finite value of emission reduction over dispersion enhancement against the additional cost of emission reduction. Unfortunately, neither the effect of sulfur oxide emissions beyond those effects on which national standards are based, nor the value of reducing those effects is quantifiable at this time. Nevertheless, such effects are real and serious. They include contribution to suspended sulfate formation, acidification of soil, streams and lakes, visibility reduction, and increase in the "background" concentration of areas adjacent to the emitting source. The factors which should be considered in assessing the necessity of dispersion enhancement for a particular source include total annual emission after control, cost of alternative control systems (including various combinations of emission reduction and dispersion enhancement), environmental elements at risk, life expectancy of the emitting source, expense which can be borne without shutdown, practice in similar industries or in industries with similar emission problems, priority for limited fuel or control technology, amenability of the source to modification, availability of land for added equipment and fuel storage, etc., any of which may create difficulties that warrant procedures to attain air quality standards by tall stacks and varying emission rates.
An added surveillance burden on control agencies is expected when dispersion enhancement is used. This is due to the fact that dispersion enhancement depends on the prediction of, and response to, continually changing meteorological conditions. It is recommended that any State choosing to allow dispersion enhancement adopt a licensing fee to cover this added surveillance expense.
The intent of these proposed regulation changes is to provide States who have large existing isolated sources of sulfur dioxide emissions another control technique for attaining national ambient air quality standards in a timely fashion and without unreasonable social disruption.
Appendix Q sets forth the conditions under which the technique may be applied, describes a comprehensive variable emission control system, defines the elements of the system and provides criteria for an acceptable regulation which authorizes the implementation, operation and enforcement of a system.
These changes are not intended to:
1. Allow the unnecessary emission of sulfur oxides into the ambient air.
2. Allow dispersion enhancement techniques to displace emission reduction techniques which are available and cost effective.
3. Allow the use of emission control methods that cannot reliably attain national air quality standards.
4. Allow the use of emission control techniques which circumvent or inhibit surveillance and enforcement of air quality standards.
5. Allow dispersion enhancement techniques in areas where there are numerous interacting sources.
6. Allow the use of dispersion enhancement techniques in or near urban areas.
7. Allow the use of dispersion enhancement techniques for pollutants other than sulfur oxides.
8. Require a State to allow dispersion enhancement techniques.
Interested persons may submit written comments on the proposed regulations in triplicate to the Office of Air Quality Planning and Standards, Environmental Protection Agency, Research Triangle Park, N.C. 27711. All relevant comments postmarked not later than 30 days after publication of this notice will be considered. The regulations, modified as the Administrator deems appropriate after consideration of comments, will be effective upon the date of their republication in the Federal Register.
This notice of proposed rule making is issued under the authority of –
REVISIONS TO PART 51, CHAPTER I, TITLE 40, CODE OF FEDERAL REGULATIONS
1. Revise the first sentence in paragraph (n), subpart 51.1 to read 51.1 Definitions
(n) "Control strategy means a combination of emission reduction and such other measures as may be necessary for the attainment and maintenance of a national standard, including, but not limited to, measures such as:
2. Add paragraph (q) to subpart 51.1 as follows: 51.1 Definitions
(q) "Dispersion enhancement" means the timing of the release of emissions to avoid meteorological conditions conducive to abnormally poor pollutant dispersion, and improvement in stack design and operation in order to increase the effective stack height. Such techniques are generally considered inferior to emission reduction for attainment of national standards, particularly primary standards, and will be acceptable only if emission reduction control technology sufficient to attain national standards in the required time is unavailable or infeasible.
The conditions for acceptability of dispersion enhancement techniques are set forth in Appendix Q, provided that Appendix Q to this part is not intended and shall not be construed to require or encourage a State to allow such dispersion enhancement techniques without due consideration of (1) the advantages of emission reduction over dispersion enhancement, (2) the availability and cost of emission reduction control technology, (3) the availability of low sulfur fuel, (4) the relative reliability of dispersion enhancement and emission reduction for achieving and maintaining national standards, (5) the relative difficulty and cost of surveying compliance with regulations governing dispersion enhancement and emission reduction methods.
3. Revise paragraph (a), subpart 51.12 to read–
51.12 Control strategy: General (a) "In any region where existing (measured or estimated) ambient levels of a pollutant exceed the levels specified by an applicable national standard, the plan shall provide for the degree of emission reduction and other measures necessary for attainment and maintenance of such national standard including the degree of emission reduction necessary to offset emission increases that can reasonably be expected to result from projected growth of population, industrial activity, motor vehicle traffic, or other factors that may cause or contribute to an increase in emissions."
4. Revise paragraphs (a), (b) and (e) of subpart 51.13 to read–
51.13 Control strategy: Sulfur oxides and particulate matter. (a) "In any region where existing or projected levels of sulfur oxides or particulate matter exceed a primary standard, the plan shall set forth a control strategy which shall be adjusted for the attainment or maintenance of such primary standard by July 1975.
(b) (1) "In any region where a secondary standard for sulfur oxides can be achieved through the application of reasonably available control technology and dispersion enhancement, "reasonable time" for attainment of such secondary standard pursuant to § 51.10(c) shall not exceed July 1975, unless the State shows that good cause exists for postponing application of such control means.
(b) (2) "In any region where application of reasonably available control technology and dispersion enhancement will not be sufficient for attainment and maintenance of such secondary standard, or where the State shows that good cause exists for postponing the application of such controls, "reasonable time" shall depend on the degree of emission reduction and other measures needed for attainment of such secondary standard and on the social, economic and technological problems involved in carrying out a control strategy adequate for attainment and maintenance of such secondary standard.
(b) (3) "In any region where the control strategy for attainment and maintenance of a secondary standard for sulfur oxides requires or results in extensive fuel switching, "reasonable time" may extend beyond July 1975, provided that the minimization of the demand for substitute fuel through the use of dual-fuel variable control systems has received serious consideration. In establishing a time for attainment of a secondary standard which the State considers reasonable, the following criteria shall be considered:
(i) The nature and prevalence of any adverse effects on the public welfare.
(ii) The value and useful life of existing combustion or control equipment which would need to be replaced as a result of the control strategy.
(iii) The availability and cost of any substitute fuel.
(iv) Other relevant social and economic impacts of the control strategy and pollutant emissions.
(b) (4) "Where the time for attainment of a secondary standard for sulfur oxides established by the State extends beyond Jan. 1, 1978, the State shall submit, after notice and public hearing, a reanalysis of the considerations specified in subparagraphs (b) (2) and (3) of this section at intervals of no more than three years from the date of plan approval by the Administrator. States shall apply reasonable interim emission reduction measures to minimize adverse welfare effects which occur at air quality levels in excess of the secondary standard.
(e) "Adequacy of control strategy.
(4) (i) "If dispersion enhancement is used as part of the control strategy, each source using this technique must be treated separately. It must be shown through a combination of diffusion modeling and air quality sampling that the emission rate control system and emission release characteristics are sufficient to insure that national standards will not be violated at any point significantly influenced by emissions from said source.
(ii) "The plan shall show that each source using dispersion enhancement to achieve national standards is sufficiently isolated from other sources so that observed and calculated pollutant concentrations in the vicinity may be attributed solely to that source. In some exceptional cases, two or more sources located in close proximity to one another may be treated as one source.
(iii) "Other conditions for the acceptability of dispersion enhancement as a control strategy for attainment of national standards are set forth in Appendix Q."
5. Add the following sentence to Appendix B, Part 3.1 at the end of the second paragraph:
Appendix B – Examples of Emission Limitations Attainable with Reasonably Available Technology.
3.1 Fuel combustion.
If these means are unavailable, infeasible, or insufficient to achieve national standards in the required time, then dispersion enhancement techniques, as described in Appendix Q may be considered.
6. Revise Appendix B, Part 3.4 last sentence to read Appendix B – Examples of Emission Limitations Attainable with Reasonably Available Technology.
3.4 Nonferrous smelters.
In such cases, less restrictive control can be coupled with restricted operations and or dispersion enhancement techniques to achieve air quality standards.
STAFF PAPER
INTERMITTENT CONTROL SYSTEMS
(Prepared by Monitoring and Data Analysis Division, Office of Air Quality Planning and Standards,
OAWP, EPA)
April 1973
INTERMITTENT CONTROL SYSTEMS (ICS)
Synopsis
1. The purpose of this paper is to analyze the alternative positions available to EPA on the acceptability of intermittent control systems (ICS) as a strategy element of state plans to protect air quality.
2. An ICS is a system designed to meet air quality standards by taking advantage of the continually changing dispersive capacity of the atmosphere. Through an ICS, emissions are curtailed during poor dispersion conditions to prevent ground-level concentrations from exceeding the standards. As dispersion conditions improve, emissions may be increased accordingly because the effluent will be dispersed through a greater volume. The emission variations are effected through such procedures as fuel switching and process rate variation. An ICS may be contrasted with constant control system (CCS), which reduce emissions by a fixed amount that is dictated by the worst expected dispersion conditions.
3. The past position of EPA has been to discourage the use of ICS because (a) it primarily relies on dispersion rather than emission reduction, (b) its reliability compared to that of CCS is questioned, and (c) enforcement of regulations that would have to accompany an ICS appears to be difficult and costly. The past position is conveyed in 37 F. R. 10845, May 31, 1972, and stated in 37 F. R. 15095, July 27, 1972. " ... At this time, it (intermittent control) is not considered an acceptable substitute for permanent control system for attaining and maintaining national standards. Experience with systems employing intermittent process curtailment indicates that although air quality is improved, violations of ambient air quality standards still occur. Additional experience with these systems may, however, in specific cases improve their reliability."
4. The EPA position on ICS is being reevaluated because (a) reliable systems are now being demonstrated and (b) constant control technology may not be available for meeting air quality standards, or may be much more costly than ICS, especially for short-term standards.
5. There are 150-200 facilities (power plants and smelters) whose operators are particularly likely to desire to employ intermittent control systems. These facilities emit one-third or more of the nationwide sulfur dioxide emissions.
6. The issue to be resolved is whether and under what circumstances a control strategy which includes ICS will be acceptable to EPA.
Discussion
1. Theory and operation of ICS.
The two basic methods of reducing ground-level pollutant concentrations are emission reduction and atmospheric dispersion. However, control systems which reduce emissions by cleansing the stack gases rely to some degree on dispersion. Pollutant concentrations in the cleansed stack gases are rarely less than the ambient air quality standards. The amount of CCS needed by a facility to attain ambient air quality standards is based in part on the expected dispersion of the facility’s effluent plume by the time it reaches ground level.
The rate of dispersion depends on meteorological conditions (wind speed, mixing height, stability). These conditions vary with time such that the peak ground-level concentration varies over several orders of magnitude even though the emission rate is constant. ICS takes advantage of this natural variation in dispersion potential by adjusting the emission rate in accordance with meteorological conditions so that ground-level pollutant concentrations do not exceed pre- selected values.
Methods of varying the emission rate may be adjustment of the plant's process rate, scheduling of high and low emitting processes to take place during the appropriate weather conditions, or varying fuel quality.
Depending on the circumstances, ICS may or may not reduce the average long-term emissions. If plant operation is curtailed during poor dispersion conditions, then it may be increased during good conditions to make up for the lost production. Average emissions would be about the same with or without ICS for this situation. If clean fuel is used to reduce emissions during poor dispersion conditions, then average emissions will be reduced somewhat. If fuel with higher sulfur content is used during good conditions, then average emissions could be greater with ICS. It must be concluded, therefore, that although ICS employs temporary emission limitation, the long-range control method is that of taking advantage of good dispersion rather than emission reduction. The elements of ICS operation are shown in Figure 1.
The estimate of present and future dispersion conditions is based on current and predicted weather conditions. The predicted conditions may be based on observed present weather in the vicinity of the source, on the informed opinion of a meteorologist who interprets the significance of the weather conditions and trends occurring over a wide area, or both. The rate of change of air quality measured by the monitoring network may also provide significant clues to the current and future dispersion conditions. The weather predictions are used as inputs to the operating model. The complexity of the model(s) will vary tremendously with the local terrain, the season, the geographical area; broadly speaking, with the local climate. The complexity is also a function of the characteristics of the source, such as the height of the stack. A tall stack generally enables the model(s) to be simpler (and more reliable) than if a source uses short and multiple stacks.
The models provide data or indications as to whether and how much to vary the emission rate.
These data, together with a firm understanding of the source’s operation form the basis for an emission control decision. If the model and meteorological predictions were perfect, this would be all that was necessary. No dispersion model or meteorological prediction is perfect, however, so feedback from the air quality monitoring network is used to check and at times to override the operating model calculation. There is a time delay between emission and concentration at a monitor so some “lead” or anticipation must be used when controlling on the basis of air quality data. This “lead” has the form of ground-level concentration thresholds somewhat below the standard to be met. When such thresholds are exceeded, emissions must be reduced regardless of the model output.
Data on actual meteorological conditions, monitor readings, emission rate, and predicted concentrations are stored, analyzed, and used to upgrade the operating model. Thus a properly operated ICS should become more reliable with use. (See Tab. 1, Theory and Operation of ICS.)
2. Reliability.
The reliability of an ICS is considered to be a technical rather than a policy matter. In application its reliability is as good as that of some presently acceptable stack gas-cleaning systems.
Further, ICS is a flexible approach. Even while being developed, an ICS possesses a capability to improve air quality once the decision is made to control emissions, although probably not reliably. Experience with the ICS should improve its reliability. A satisfactory level of reliability (i.e., a dependable model) might be expected in 1-2 years. Reliability should eventually approach the reliability of stack gas-cleaning methods by the time such methods could be installed (also 1-2 years). Thus, some benefits are derived from ICS during its development period; none are expected from a CCS until it is placed "on-line."
If dispersion conditions are less favorable during a season or year than expected (based on long term data), the model may be modified or the criteria for control made more stringent. If the necessary efficiency of an installed CCS were determined using weather data collected during an anomalous period, considerable delay and substantial costs may be involved in rectifying the circumstances and enabling the standards to be attained. An ICS system can react to such situations more promptly. Its flexibility enables the operator to respond to the need to attain standards within hours or days rather than months.
Nevertheless, the preferred procedure to attain National Ambient Air Quality Standards (NAAQS) everywhere, always, is to limit emissions on a continuous basis. With a combustion source the most reliable tactic is to burn clean fuels. With a process source, the most reliable tactic is to limit the rate of operation to the level that emissions pose no threat to the NAAQS during the most adverse atmospheric conditions, taking due care that the conditions on which the rate is based are, in fact, most adverse. These approaches are uneconomical for some facilities.
As a consequence, control devices to clean the stack gases are employed on many facilities. Unfortunately such devices do not operate continuously at design efficiency. EPA engineers estimate that an SO2 flue gas-cleaning device will be inoperative about 15% of the time; 5% for scheduled maintenance and 10% because of malfunctions. The threat to NAAQS created by such outages varies among facilities.
Factors such as these determine the effect of the periods of inoperation: Are the breakdowns weather related, systematic or random? Does the facility operate continuously? Does it terminate operations when the devices are inoperative? If an uncontrolled facility threatens the standards 75 days per year, and control device malfunctions are random, the NAAQS would be violated on about 11 days per year or 3 % of the time.
Similarly, if the operators of an ICS err not more than 15% of the time (15% x 76 poor dispersion days – 11 days), the ICS will protect the NAAQS as effectively as a CCS.
Data on the effectiveness of ICS are sparse. TVA reports that when the decision to curtail operations was made 18 hours before a curtailment was required to be initiated, 18 of the decisions were in error. However, additional updating procedures are now used and the "go-no go" decision is executed 2 hours before curtailment is required.
Data indicating the effectiveness of ICS systems are available from TVA, the Puget Sound Air Pollution Control Agency and ASARCO.
The TVA and ASARCO data are reported by them. TVA data are the more objective because a date for inaugurating the ICS was established (Sept. 1969). The ASARCO data indicate an increasing capability to reduce violations of air quality standards at the sites where air quality is monitored by an ICS program. ASARCO has operated ICS programs since 1969 at increasing levels of effort. The Puget Sound APCA data, though based on local standards, indicate an improved reliability of the Tacoma ICS operation.
In summary, an ICS, when properly designed and diligently and conscientiously operated, can be used to attain air quality standards with the same reliability as a CCS. (See Tab 2., Reliability of ICS and CCS.)
3. Enforcement.
For an enforcement system to succeed it must provide an adequate incentive for sources to comply with emission and air quality regulations. Adequate incentive exists if the regulation associated with the ICS (1) provides for adequate control agency surveillance of the source and its impact on air quality, (2) enables the agency to establish liability if air quality or emission standards are violated, and (3) prescribes sufficient penalties to deter a source from allowing such violations to occur.
If the incentives are adequately provided for, the control agency has four basic approaches to enforcement of an ICS:
1. Enforcement on air quality. The source operates the ICS and is held directly responsible for maintaining air quality standards in vicinity of the plant.
2. Enforcement on emissions. The source operates the ICS and is required to vary emissions in accordance with pre-arranged "curtailment criteria." These criteria are specific meteorological conditions or air quality levels at which the source curtails emissions by predetermined amounts.
3. Enforcement on emissions and air quality. This is a combination of the preceding approaches. The source basically operates in accordance with curtailment criteria, but simultaneously is responsible for maintaining air quality.
4. Enforcement by control agency operation of ICS. The agency, on an operational basis, determines when and in what manner the source varies emissions to attain and maintain air quality standards.
When air quality is the basis for enforcement (Approach 1), the source, as a condition for being permitted to use ICS, assumes full responsibility for maintaining the air quality standards. To assure that the standards are maintained, the control agency has access to air quality data on a real-time basis and has access to all air quality sensors to assure that they are operated, maintained, and calibrated properly. Enforcement actions are initiated if air quality standards (or regulations) are exceeded. This approach allows the source the maximum degree of flexibility.
When emissions are the basis for enforcement (Approach 2), the source, as a candidate for being permitted to use ICS, provides the control agency with a set of curtailment criteria. These are meteorological conditions (and occasionally air quality levels) which, in the course of developing the ICS, have been ascertained to be precursors or indicators of the need to limit emissions to avoid air quality violations. The agency requires access to air quality, meteorological and emission data. Enforcement actions are initiated if the source does not properly adjust emissions when conditions meet the curtailment criteria. This approach has the advantage of requiring curtailment even though an air quality sensor is not located in an area where the ground-level contamination is most likely to exceed the air quality standard. It does not require the source to assume responsibility for maintaining air quality standards. Repeated violations of the air quality standards would be corrected only by periodic reviews of the system by the source and agency, at which time revision of the "curtailment criteria" would be in order.
Enforcement on emissions with responsibility to maintain air quality standards (Approach 3) is a combination of the preceding approaches. "Curtailment criteria" are developed and justified to the control agency. Nevertheless, the source is immediately responsible for any violations of air quality standards. This approach entails continuous monitoring of emissions, dispersion conditions and ambient air quality by the control agency. If properly operated, it protects against violations of air quality standards in areas where no sensors are located; against unquantified effects of pollutants (see Tab 4); and provides prompt feed-back to improve the curtailment criteria when the air quality data show the criteria to be inadequate.
Enforcement by control agency operation of the ICS (Approach 4), in essence, requires the agency to operate the facility. The source has no flexibility. It responds to the direction of the agency. This approach is likely too paternalistic and so philosophically divergent from normal economic and industrial practices as to be unacceptable to any source.
Any of the four approaches, due to the relative complexity of an ICS would impose a considerable administrative, surveillance and enforcement burden on a control agency. The burden is compounded if sources are located in rugged terrain, if more than one source is involved, or if sources are not isolated from each other. Particularly troublesome is a multi- source system or system operated where the background levels of contamination exist. Under these circumstances establishing liability for violations is difficult, uncertain and time consuming.
Further, EPA and most State and local control agencies are not staffed to cope with the burden of enforcing the requirements of the CAA in areas where several intermittent control systems are present. If use of ICS is not carefully restricted, the enforcement burden can easily become unmanageable.
A reasonable remedy to the cost of enforcement to the agency would be to require a permit to operate an ICS. The permit fee would be set at a level which would pay for the costs of surveillance and enforcement of the system.
In summary, the problem of enforcing an ICS is a major reason for the reluctance of many to endorse the use of such systems. If ICS is limited to use by single, isolated sources, enforcement appears manageable. If allowed to be applied by multi-sources in urban areas, enforcement requirements place great demands on the resources of air pollution control agencies, including those of the EPA. Necessary resources might be acquired from fees for permits to use an ICS. Assuming administrative problems are overcome, the preferred approach is to enforce on the basis of emissions and air quality. (See Tab 3., Enforcement.)
4. Legal Position of ICS.
The most important policy decision relating to ICS (and tall stacks) is the interpretation of Section 110(a)(2)(B) of the Clean Air Act. This section requires that the SIPs achieve NAAQS through "... emission limitations ... and such other measures as may be necessary ... This key phrase may be interpreted in two ways. It may be construed to mean that "other measures" may be used only if sufficient emission limitation means are unavailable or infeasible, thus making "other measures" necessary. Or it may be interpreted to mean that any combination of emission limits and "other measures" may be used, provided NAAQS are attained.
Both ICS and tall stacks are "other measures." Both techniques rely on the dispersion of pollutant emissions through a larger volume of air to reduce ground-level concentration. A taller stack does not reduce emissions. ICS reduces emission sometimes, but may increase emissions at other times. The average emissions are reduced only slightly if at all.
If ICS and tall stacks are to be rejected in favor of the more costly emission limitation methods (CCS), and this cannot be done on reliability grounds, then EPA must adopt and defend the interpretation of Sec. 110 that "other measures" may be used only when emission limitation is unavailable and/or infeasible. (See Tab 4., Legal Position of ICS.)
5. Unquantified Effects.
The present ambient air quality standards are first steps towards quantified standards of environmental quality. However, they do not yet include such effects as contribution to background concentration, conversion of SOx to suspend sulfates, acid rain, climatic change, and long-range ecological damage. All these presently unquantifiable effects will be reduced if emissions are limited but not if NAAQS are attained solely by dispersion of the contaminants.
Unquantified effects may also include restrictions on growth, particularly in the vicinity of large point sources which operate an ICS. If the objective is only to attain standards in the vicinity of the source, then the air quality will have been usurped and no other additional facility may be located in the neighborhood of the source. (See Tab 5., Unquantified Effects of Pollutant Emissions.)
6. ICS and the SIPs.
The state implementation plans (SIPs) and EPA procedures for determining their acceptability are primarily based on air quality control through emission reduction. Acceptance of dispersion techniques (viz., ICS) in lieu of emission reduction in more than a carefully limited number of cases may require major revisions in the SIPs. Unless considerable care is taken in defining and limiting the situations in which EPA will accept ICS as a control measure, the basic SIP philosophy of control through emission reduction may be undermined.
7. Costs of ICS.
An ICS is not cheap. The costs occur in three areas: Direct installation and operating costs to the source, lost production for the source and lost wages for its employees, and costs for surveillance and enforcement to the public sector.
The direct costs to the source include equipment to monitor emissions, weather and air quality; computer and modeling services; additional technical and scientific personnel; etc. It frequently requires one-year and $300,000 to $400,000 to develop an elementary ICS system and another $100,000 to $150,000 to maintain and operate it.
Lost production may or may not be a serious cost factor depending on whether the method of emission reduction is production curtailment, whether the facility operates at full capacity, whether lost output can be recovered without serious cost penalties, and on lead time for curtailment. ASARCO indicates that they curtailed annual production 30% at one plant. EPA estimates the curtailment may have cost $800,000 to $1,000,000.
The increased cost of surveillance and enforcement may be considerable. It is reasonable to expect the source to defray at least a part of this public expense.
The relative cost of ICS and alternative CCS will vary widely with the particular conditions. TVA data indicates that limestone scrubbers might be 10 times as costly as ICS for meeting short-term SO2 standards near TVA's power plants. Kennecott estimates that the cost of 90% reduction of sulfur emission by CCS would cost 50% more at their Utah smelter and 350% more at their Nevada smelter than a least cost system which employs a combination of CCS and ICS. (See Tab 6., Cost Effectiveness.)
8. Estimate of Number of Facilities Involved.
There are 379 coal-fired power plants in the U. S., each of which consume more than 50,000 tons of coal annually. About 85-100 of these are located in remote or rural areas and are required by state regulations (as indicated in the SIPS) to reduce the sulfur content of their fuel to 1 % or less. These plants currently emit about 13% of the sulfur dioxide emitted nationwide. If the critical sulfur content of the fuel is 2%, the number of such facilities rises to 150-200. Nationwide they emit about 26-28% of the Nation's sulfur dioxide.
Most of the country's 16 copper smelters, because of their sites, the magnitude of their emissions and their inability to control their sulfur dioxide emissions sufficiently, threaten the NAAQS. Currently, they emit 3.5 million tons of SO2 annually.
In summary, 100-115 facilities probably will apply for permission to meet SO2 NAAQS by ICS. Another 85-100 facilities, making a total of about 200, are likely to apply because of state limitations on the sulfur content of the fuel. These 200 plants emit one-third or more of the nationwide emissions of sulfur dioxide. (See Tab 7., Number of Facilities which may Employ ICS.)
9. Self-Retirement Factor.
An Intermittent Control System, though currently cost-effective in many situations, over a period tends to be self-retiring. As constant control systems increase in reliability, and decrease in costs, the differences between the ICS and CCS cost-benefits decrease. The ICS is an inconvenience to the operator. Costs of idled workers, equipment, stock piling of raw or partially processed materials add a "harassment" aspect to the operation of the ICS. It is reasonable to anticipate that it eventually will become less attractive as a control tactic for many sources and some source categories.
10. Pollutant.
The availability and cost-effectiveness of control methods vary with the pollutant. Relatively inexpensive and very efficient CCS is available to control particulate matter emissions from point sources. Similarly, cost-effective controls for CO from stationary sources are available.
Therefore, ICS is not necessary for the control of these pollutants. The very high percentage of NOx and HC emissions from mobile sources and atmospheric chemistry considerations make these two pollutants poor candidates for ICS. On the other hand, control techniques for SOx emitted from combustion sources are expensive, supplies of low sulfur fuel are inadequate, and methods to cleanse SO2 from exhaust gases are not very efficient. Non-ferrous smelters emit greater amounts of SO2, much of which can be captured at reasonable costs. However, in some situations and in some locations, sufficient amounts of SO2 would still be emitted to threaten NAAQS.
In summary, of these pollutants, ICS is warranted as a control tactic only for SO2.
11. Source Size.
The cost of monitors, modeling and enforcement limit the use of ICS to large sources. Only if these services are provided by the control agency will it be feasible for small sources to use an ICS.
12. Source Isolation.
The contribution of a source's emissions to the observed ground-level concentration is a vital factor in the development, operation, upgrading and enforcement of an ICS (see Fig. 1). If a source is sufficiently isolated, and the pollutant of concern is SO2, it can be assumed that all of the observed ground-level concentrations in the vicinity of the source are due to the source. The development, improvement and enforcement of the ICS can be straightforward, unambiguous, and reasonably objective. If a group of sources is isolated and a prior agreement can be obtained on the division of liability among the sources, then such isolated clusters may be treated as one isolated source. Sources located in or near urban areas, where some contribution to the ground- level concentration may result from several or many smaller sources, pose severe technical and enforcement difficulties.
The use of an ICS should be limited to isolated sources of SO2. Otherwise, difficulties in developing, upgrading, enforcing and assessing liabilities may be created which increase costs and jeopardize attainment and enforcement of the NAAQS. A criteria for isolation might be that an ICS be approved for a source only if it were located in an area where contributions from other sources to ground-level contamination do not exceed 10% of the annual NAAQS.
Options for acceptability of ICS
There are three broad options as to the acceptability of ICS for existing sources: always, never, and sometimes.
In all cases the option refers to a proposed use of ICS that is technically capable of meeting air quality standards with a reliability equivalent to acceptable CCS, legally enforceable, and acceptable to the state agency. These technical and enforcement conditions will severely limit the number of proposals for ICS use.
The policy options cover the likely cases where ICS would be economically attractive to a source and could meet EPA requirements for reliability and enforcement. Such cases would primarily be large, non-urban sources of SO2.
Option 1: Accept ICS for any existing source of pollutant if the proposed control system is technically sound and legally enforceable for meeting AAQS.
Pro:
(a) The cost of meeting NAAQS will be lowered.
(b) The demand for low sulfur fuel will be lowered.
(c) Large sources will be able to respond flexibly to changes in NAAQS and to extreme meteorological conditions.
(d) NAAQS will be attained sooner.
(e) Legal support for this option may be found in Sec, 110 (a) (2) (B) of the Clean Air Act in the words "... and such other measures as may be necessary to insure attainment and maintenance of such primary or secondary standard ..."
Con:
(a) This option is the most difficult to legally defend.
(b) Annual emissions from sources using ICS would not necessarily be reduced and may even increase in some cases.
(c) The SIP preparation and approval process will be seriously upset by this option.
(d) Both growth and degradation problems will be increased.
(e) State and federal resources for plan evaluation and enforcement will be strained.
(f) The potential benefits of ICS may not be realized due to over-taxed surveillance resources.
(g) Available and economically reasonable CCS will not be used in some cases due to the lower cost of ICS. This is especially true of particulates.
Option 2: Reject ICS. Accept only CCS. Allow compliance delay until 1977 if necessary for development of CCS.,
Pro:
(a) This option is the most legally defensible.
(b) This is consistent with past policy that NAAQS are to be met by permanent emission reduction.
(c) ICS may be used under this option as an interim control measure in accordance with Sec. 110(f) (1) (c).
(d) No revision of SIPs is required.
(e) No extra burden is placed on surveillance and enforcement resources.
(f) Pollutant emissions are minimized.
Con:
(a) The cost of meeting NAAQS will be highest under this option.
(b) The demand for low sulfur fuel will remain high.
(c) Little flexibility will be available to sources to respond to changes in air quality standards or exceptionally poor dispersion.
(d) Attainment of NAAQS will be delayed.
(e) The position of EPA will be inflexible. In some instances prohibitively expensive CCS or permanent production curtailment will be the only available control options.
Option 3: Accept ICS only under certain conditions.
Discussion: Three suboptions are presented below. Each suboption is a method of separating acceptable from unacceptable uses of ICS in addition to the requirements that ICS be reliable and enforceable.
Suboption 3a:
Require available control technology to be applied. Allow ICS if available technology is insufficient to achieve air quality standards. Require replacement of ICS by CCS when new technology becomes available.
Pro:
(a) Minimum deviation from past policy.
(b) Relatively easy to defend legally. ICS may be used when necessary to meet NAAQS (Sec. 110(a) (2) (b)).
(c) Minimum revisions of SIPs required.
(d) Minimum burden on surveillance resources.
(e) Minimum relaxation of emission reduction requirements.
(f) An alternative to permanent plant operation curtailment is available if the CCS is insufficient to achieve air quality standards.
Con:
(a) Permanent controls may be technically available but economically impossible. In such cases EPA would be put in the position of forcing partial or total plant operation curtailment while withholding an effective and economically viable control method.
(b) ICS may be much more cost-effective than any CCS, particularly where short-term NAAQS are violated only a few days or hours per year. EPA would be put in a position of defending an economically irrational policy.
Suboption 3b:
Allow ICS only for attainment of secondary NAAQS. Require COS for the attainment and maintenance of primary NAAQS.
Pro:
(a) ICS is more subjective in design than CCS and in some cases may be less reliable than the best CCS. Its use should be prevented when health-related standards are involved.
(b) This option provides a clear-cut criteria for the acceptance or rejection of ICS.
Con:
(a) ICS, when properly designed and conscientiously operated, can be as reliable as stack-gas cleaning methods which are acceptable for the attainment of primary standards.
(b) The cost of meeting NAAQS is high under this option.
(c) The demand for low sulfur fuel will remain high.
(d) Attainment of NAAQS may be delayed.
(e) The position of EPA will be inflexible. In some instances prohibitively expensive CCS or permanent production curtailment will be the only available control options for meeting the primary standard.
Suboption 3c:
Determine the acceptability of ICS on a case-by-case basis. Base the decision on the availability and cost and expected emissions of alternative control systems, the expected life of the plant, the frequency and severity of pollution due to the plant and any other relevant factors. Review the decision periodically. Require CCS when conditions change in its favor.
Pro:
(a) Allows for the optimization of public benefits by balancing the value of emission reduction against the cost of such reduction.
(b) Allows the use of a complete range of emission reduction and dispersion techniques for the timely, effective, and economical improvement of air quality.
(c) Avoids unreasonable decisions due to inflexible criteria (i.e., allowance of ICS for particulates when cost-effective CCS is available or the requirement that multi-million dollar SO2 scrubbers be used to avoid violations expected only a few days per year).
(d) Allows the States maximum flexibility for meeting NAAQS on time and at reasonable cost.
Con:
(a) Case-by-case decisions may result in case-by-case lawsuits – at least until a precedent is established.
(b) The lack of objective criteria may lead to charges of arbitrariness, inequity, or favoritism.
(c) An accurate, quantitative measure of the environmental effectiveness of emission reduction is not available for use in optimizing the cost-effectiveness of emission control.
(d) The acceptance procedure will be inherently lengthy and complex.
(e) Negotiations over the acceptability of ICS may delay the application of any control.
Summary arguments
A basic premise to the recommended EPA policy on ICS is that there are benefits to society from reducing emissions of most pollutants into the atmosphere independent of the attainment of NAAQS. An important portion of the effects of pollutant emission is presently unquantifiable (e.g., acid rain, corrosion, suspended sulfates, property damage, ecological change). These effects are a function of atmospheric loading and are reduced most effectively by attaining NAAQS through permanent emission reduction. They are not alleviated through use of tall stacks and slightly alleviated by ICS, if at all.
The benefits of emission reduction over emission dispersion are of finite, not unlimited, value. The outright rejection of dispersion techniques when emission reduction techniques are unavailable or prohibitively expensive, would not optimize benefits to society. It would be unreasonable to force permanent curtailment of production or severe increases in product prices due to the insistence on CCS if reliable and much more cost-effective ICS is available.
The problem has been to find a decision-making framework to determine an acceptable strategy for attaining NAAQS which optimizes benefits to society. Many alternatives are available to accomplish this objective, from unrestricted use of any tactic that will meet NAAQS to the prohibition of ICS or tall stacks. Neither of these extreme alternatives allows the optimization of benefits.
Combination alternatives that allow ICS only to solve part of the problem (i.e., after application of reasonably available, or best, or most practicable control technology, or after attaining primary NAAQS by permanent emission reduction) do not really address the problem of maximizing benefits. They have the advantage of providing more dogmatic criteria for decisions on acceptable strategies. However, they are arbitrary in initial selection of criteria and once selected they are relatively inflexible. Such inflexibility can force unreasonable decisions in some circumstances for, in fact, there are few situations which are precisely alike. The cost advantage of ICS over CCS is 1:10 or more for power plants with tall stacks where short-time SO2 standards are violated only 1-2 % of the time. Some smelters see no advantage in attaining standards by ICS alone. A combined CCS-ICS system may be less costly than either CCS or ICS.
The cost advantage of a combined system over CCS ranges from 1:1.5-1:3.5 depending on the smelter involved. Other variable factors include the local climate and topography, the expected life of the plant, population and biota surrounding the plant, etc.
It appears logical that the way to make the proper decision in each case is for EPA or the State to decide each case individually within a consistent policy framework. It is essential that the framework formally recognize that the growth of population and economy, and changes in technology create increasing stress on the environment; recognize that there are benefits of emission reduction over dispersion at the present, as well as in the future, and recognize that such reduction has a finite, not unlimited, value that must be compared to the cost differential between emission reduction and dispersion.
This policy would lead to many generalizations that predetermine the decision in most cases. For example, ICS would be limited to attaining SO2 standards, would apply to isolated sources, would consider terrain problems, would apply to sources whose processes are adaptable to variable levels of operation, would apply particularly to facilities which threaten short-term standards, etc.
This policy allows States maximum flexibility to devise acceptable control strategies for attainment of NAAQS. If they can place a high value on the benefits of generally clean air, they can require much emission reduction; EPA has not "sold them out." (EPA should continuously and vigorously support States in their desire to achieve highest quality air attainable.) If the State wishes to use ICS to meet NAAQS, they can allow it and, with reasonable justification, EPA can approve the strategy. Since decisions will be made on the facts in each case, EPA can avoid being forced into unreasonable actions.
In summary, Camp recommends the following three-tiered approach to ICS:
1. Adopt a policy that emission reduction is preferred to dispersion techniques even though NAAQS can be reliably attained in some circumstances by the latter techniques.
2. Decide whether a source is a viable candidate for ICS on a case-by-case basis, taking into account the availability and cost of CCS, and the numerous other factors that may be relevant to the particular case.
3. Scrutinize the reliability and enforceability of the proposed ICS, if it is deemed feasible for the source to use ICS as a part of its control strategy. The ICS, if approved, would be reevaluated periodically (1) to determine if cost-effective CCS had become available, and (2) to determine if the reliability of the ICS is adequate and improvable. (See Tab 8., Conditions for Acceptability of an ICS.)
TAB. 1. THEORY AND OPERATION OF ICS
Definition
In the broadest sense an Intermittent Control System (ICS) is the deliberate variation of pollutant emission rate based on estimates of atmospheric dispersion potential order to reduce the environmental impact of those emissions.
This broad definition includes many practices not generally considered to be intermittent control, such as soot-blowing or agricultural burning during good dispersion conditions and the alteration of work schedules to change the time and intensity of peak traffic density. A narrower and more familiar definition of ICS refers only to emissions from stationary sources which operate continuously or during fixed working hours and which reduce their emission rate during periods of poor atmospheric dispersion conditions in order to avoid the high ground-level pollutant concentrations probable under those conditions.
The second, narrower definition will suffice for our purposes provided two qualifications are added: (1) The emission rate may increase during periods of good dispersion as well as decrease during poor conditions, and (2) ICS is not necessarily an exclusive control strategy; it is one of several control strategy elements which may be used in combination to minimize the impact of pollutant emissions on the environment. These control strategy elements include process change, removal of pollutants from the exhaust stream, and the use of cleaner fuel, all of which reduce pollutant emissions and ICS, taller stacks, and appropriate siting, which rely on improved dispersion of pollutant emission to reduce or redistribute the ground-level pollutant concentration.
Variable dispersive capacity
It has been observed that pollutant concentrations resulting from a constant rate of emission vary over several orders of magnitude at any given ground-level monitor. This phenomenon is due to three mechanisms, all related to temporal variation in meteorological conditions : (1) The pollutant is transported toward or away from the receptor due to changes in wind direction, (2) the maximum ground-level concentration attributable to a source varies both in magnitude and distance from the source with changes in atmospheric stability and wind speed, and (3) the pollutant is mixed throughout a larger or smaller volume of air due to changes in wind speed, stability and mixing height.
The first mechanism is of limited interest for it does not necessarily reduce the ground-level concentration, but only moves that concentration from place to place. The other mechanisms are of considerable environmental significance.
The adverse effect of a pollutant on an environmental element (animal, plant, material) is an increasing function of the rate of transfer of that pollutant to that element. This rate of transfer is, in turn, an increasing function of the atmospheric concentration of the pollutant to which the element is subjected. The relationship between pollutant concentration and adverse effect is most probably non-linear. At high pollutant concentrations, the effects may be rapid and intense; at lower average concentrations the effects may be slow, subtle and cumulative; at still lower average concentrations, there may be no adverse effect; in fact, some waste products may have a net beneficial effect on some environmental elements when present in low concentration.
The relationship between pollutant emission rate and ground-level concentration of that pollutant is, therefore, of considerable importance. If a pollutant emitted into the atmosphere is dispersed through a greater volume of air, then its impact on the surface environment will be less per unit of surface area, but more surface area will be affected. If all the pollutant is eventually deposited on the surface, then the product of (pollutant per surface area) times (surface area effected) is constant. If, however, the adverse environmental effect due to ground-level concentration and pollutant deposited on the surface decreases more rapidly than pollutant per unit volume and pollutant per unit area, then wider pollutant dispersion will reduce total environmental impact of those pollutants. NAAQS are based on the assumption that no adverse effect occurs below a threshold ground-level concentration. This assumption is undoubtedly simplistic (See Tab. 4: Unquantified Effects of Pollutant Emissions). However, the assumption that total environmental damage decreases with increased pollutant dispersion is reasonable. It may be concluded, therefore, that the capacity of the environment to absorb waste with a given environmental impact (not necessarily at zero impact) varies with the dispersion of that waste throughout the air and, eventually, over the Earth's surface.
Effectiveness of ICS
Now consider two emission sources which produce the same long-term average amount of emissions under identical circumstances except that one source emits at a constant rate and the other source varies its emission rate with dispersion conditions. Both the peak and average environmental impact of the source using ICS will be less than that of the source that emits at a constant rate. (See the appendix for an example supporting this statement.) If a source using ICS
has a lower average emission rate than a constant rate source (in addition to variable emission rate based on dispersion potential), then its environmental impact will be that much less. Thus, there are at least two cases in which ICS is clearly environmentally superior to CCS.
The third, and most difficult case, is where the ICS source emits more pollutant on the average than the CCS source. There is certainly a point at which the greater emission from ICS is no longer compensated by the wider distribution of that emission. The determination of this balance point depends on the indicator of environmental impact used, the accuracy of dispersive capacity prediction and the relationship between emission rate and dispersive capacity. It is sufficient for our purposes here to state that the relative environmental effectiveness of ICS in comparison with CCS becomes positive at some point where average ICS emissions are somewhat greater than average CCS emissions, and increases as ICS emissions are reduced. This statement is illustrated in Figure 1-1.
ICS operation
There are three types of ICS. These are not mutually exclusive. In fact, an ideal ICS would include all three.
1. Open loop system based on diffusion modeling.
This system is illustrated in Figure 1-2a. The heart of this ICS is the operating model. This is a diffusion model that estimates maximum ground-level concentrations based on emission rate, meteorological conditions (wind direction, wind speed, stability, mixing height), and local topography. The meteorological conditions needed to operate the model generally must be predicted. The prediction will be based on the short- and long-range past history of the weather in the vicinity of the plant and on National and regional weather forecasts. The desired or expected plant emission rate is also entered in the operating model.
One form of model output is the expected maximum ground-level concentration. This estimate is compared with the relevant short-term air quality standard. If the expected maximum concentration is less than the standard (with some safety factor included to compensate for uncertainty), then no action is taken and the plant operates normally. If the estimated maximum concentration exceeds the appropriate threshold, then plant emission must be reduced. Emission reduction may be achieved by switching to cleaner fuel, reducing the level of plant operation, or delaying high emission processes that may have been scheduled.
2. Closed loop system based on air quality monitoring. (Figure 1-2b)
This ICS relies entirely on real time air quality feedback for information on which to base the control decision. An array of continuous air quality monitors is located at those points where maximum concentrations are expected. The level and rate of change of pollutant concentration at each monitor is continuously scanned. The emission rate is curtailed whenever a monitor indicates that a standard is in danger of being exceeded. Because of the time delay involved in reducing the emission rate and because of the time required for the pollutant to travel from the stack to a monitor, control must be initiated somewhat before monitored concentration reaches the standard. Threshold values of concentration and rate of change of concentration will be set, based on the reduction response time and source-receptor distance. The amount of reduction needed is not estimated by this system. A step-wise reduction schedule would be appropriate.
3. Closed loop system based on diffusion modeling upgraded by emission-concentration data (Figure 1-2c)
This system is similar to the first except that air quality monitors, data storage, and periodic upgrading of the operating model have been added. In most instances, a detailed climatological study and a validated diffusion model will not be available at the initiation of an ICS. The collection of emission and concentration data provides a basis for analysis of the model's accuracy, and for possible improvements in that accuracy. The monitoring network required would be similar to that discussed under an ICS system based solely on air quality monitoring (2, above), except that fewer monitors would be required because the model is available to interpolate air quality between monitors.
The existence of a monitoring network may improve meteorological prediction, especially in cases of complex topography, because source-receptor pollutant transport is an indicator of meteorological conditions.
4. Combined system (Figure 1-2d)
An ideal ICS would employ three complementary operations. The emission source would first look at meteorological predictions and adjust its emission output accordingly. If that adjustment were not sufficient, as indicated by air quality monitoring, then additional procedures for further emission reduction would be activated. Records of the emissions, measured concentrations, and meteorological conditions would be continually or periodically analyzed to determine if improvements in the prediction accuracy of the model could be effected. Such improvements would then be incorporated in the model.
The advantage of a combined system lies in the fact that each of the loops performs a different, valuable function. The model allows lead time in performing control operations. This is desirable in that time is necessary to switch fuel or curtail operations. Furthermore, there is a lag between the emission of pollutants and the registration of their effects at the monitors. Even if all control functions were instantaneous, this lag time could still result in unacceptably high concentrations at the monitoring site(s) for a limited period. The lead time associated with the model is a valuable compensation for the several system lags. The air quality loop firmly establishes the connection between air quality and emissions. During the initial operation of a combined system, when the operating model is tentative and the dependency of air quality on local meteorology only partially known, it is possible that the air quality loop would often be the controlling one. As time goes on and data are accumulated, it should be possible to improve the operating model so that the air quality loop is activated less and less frequently and the overall ICS operation becomes smoother and more predictable.
APPENDIX TO TAB. 1.
COMPARISON OF ICS AND CCS FOR EQUAL AVERAGE EMISSION
The example below is for illustrative purposes only. It does not represent an actual or proposed control system.
[Table Omitted]
Column (1) represents the frequency that an ambient ground-level air quality concentration represented in column (2) as a result of a source's operation. Column (3) indicates that 50% constant control is applied. Column (4) indicates the concentration expected as a result of the constant control. Column (5) represents the amount of ICS control applied to each frequency of occurrence interval. Column (6) is the expected ambient concentration as a result of the degrees of ICS control applied during each frequency of occurrence interval.
Not only is the peak concentration reduced when ICS is used, but the average concentration is reduced as well. The reader is invited to substitute any schedule of radiation which averages to 50% control and decreases from top to bottom for the schedule used in column (5) to assure himself that both peak and average expected concentration will be less for ICS.
TAB 2. RELIABILITY OF ICS AND CCS
Many control officials are reluctant to accept an ICS as a control measure because the reliability of such systems for protecting national ambient air quality systems (NAAQS) has not been adequately demonstrated. Recent data have become available that permit a judgment to be made of the reliability of an ICS.
The traditional and preferred procedure to attain and maintain in the ambient air quality standard is to limit emissions on a continuous basis to the extent that NAAQS are not exceeded during the most adverse meteorological conditions. With a combustion source, the most reliable technique is to use fuels which contain sufficiently small amounts of the polluting elements and to use them in a manner such that pollutant emissions are kept to a minimum. Where a process source is the threat to NAAQS, the most reliable technique is to maintain a rate of operation such that emissions are sufficiently small to constitute no threat to NAAQS. For some facilities these approaches impose serious economic consequences.
Control devices to clean the exhaust gases have been (and are being) developed to limit emissions to comply with regulations designed to attain NAAQS. Unfortunately, such devices often do not operate continuously at design efficiency. EPA engineers estimate that SO2 flue gas cleaning devices will be inoperative for scheduled maintenance at least 2 weeks per year and for unscheduled repair an additional 10% of the time.
The threat to the NAAQS created by such outages varies among facilities, depending upon whether the breakdowns are systematic or random; whether the facility operates a 24-hour day, 7-day week; whether the facility terminates operations when the control devices are inoperative; whether the malfunctions are weather related; etc..
Let us assume that a continuously operating facility without control devices causes NAAQS to be exceeded on 20% of the days of the year; that its control devices, which when operating are sufficient to eliminate threats to NAAQS, are inoperative 15% of the time; and that the facility operates at normal capacity whether or not the control devices are operating. Then, if malfunctions of the control devices are random, the NAAQS would be expected to be exceeded on 3% of the days (15% x 20,1,1.=3%) or 11 days per year.
Let us now assume that the facility operates an ICS to curtail emissions during periods when NAAQS are most in jeopardy and that the threats occur on 75 days per year. It would be required that the operators of the system err in the direction of too little control on not more than 15% of these 75 days for the ICS to protect NAAQS as effectively as the control devices. This is a reasonable and attainable standard for reliability of an acceptable ICS.
Information on the effectiveness of ICS as a procedure to protect air quality is limited to data from operators of the systems, especially TVA (at the Paradise Steam Plant) and ASARCO (at the Tacoma and El Paso smelters). Furthermore, the indicated effectiveness of the system may be closely tied to the number of air quality sensing sites if the objective of the operators is primarily to avoid violations of the standard at the sampling sites. (It must be pointed out that an ICS system went into operation at the Trail, B. C., smelter in the early 1940's. TVA operated a system for a period in the middle 1950's at their Hinston Steam Plants However, comparative data are not readily available for before and after implementation of the ICS.)
TVA reports the following "before and after" data for their ICS at the Paradise Steam Plant.
These data are from 14 sensing sites within a 221/2 degree sector centered on the 331/2 degree azimuth from the source. Approximately 10% of the wind directions cause the plume from the plant to threaten this sector. TVA curtails emissions without regard to wind direction so it is expected that unsensed violations would be not more than 10 times those reported. Since weather situations which are conducive to high ground-level concentrations occur more frequently when winds have a southerly than a northerly component, the factor of 10 alluded to is undoubtedly a maximum.
ASARCO reports the following numbers of violations in vicinity of the El Paso smelter to the variance to the Texas SO2 standard (0.5 ppm for 1-hr.)
Year and number of violations
1970 100
1971 26
1972 30
These data are based on data sensed at 18 sites. The system incorporates a continuous feedback of air quality information to the control center. Therefore, the operators have information as to impending threats to the standards at the sites.
ASARCO independently reported that the 24-hr S02 NAAQS was violated 3 times near El Paso in 1970 "when the fully telemetered closed-loop system became operational" and 2 times in 1971. The same source reported the 3-hr SO2 NAAQS was violated 8 times in 1970 and 2 times in 1971.
The following air quality trends near their Tacoma, Washington, smelter are reported by ASARCO 4 and Puget Sound APCA.
These data are not strictly comparable for a number of reasons: The facilities are in different climatic and topographic situations; the systems are devised to meet different standards; and one is a combustion source, the others, process sources. Nevertheless, substantial reductions in pollutant levels occurred at the sensing sites in all cases. Where 24-hour data are available, the evidence is strong that the 24-hour primary SO2 standard may be attained near large isolated point sources by ICS methods. Violations of shorter-term standards are reduced by a factor of 4 to 5.
Several caveats are in order:
a. The reliability of an ICS is a function of the vigor with which the system and standards are policed and enforced. ASARCO for example, established their systems first at facilities which were near populated areas. Public concern provides an incentive to attaining the standards.
b. The indicated reliability of an ICS may be a function of the number and placement of air quality sensors. Data from Texas APCS 2 suggests that the hours of violation increased roughly linearly from 1968 to 1970 with the increase in the number of monitoring stations.
c. A well-operated ICS may be expected to become more reliable with time. The operators acquire experience and a better understanding of the nature of their problem. On the other hand a CCS may decrease in reliability with time due to aging and wear.
In conclusion, for some sources an ICS may as effectively protect against violations of the NAAQS at ground level as a CCS.
REFERENCES
[Footnotes Omitted]
TAB 3. ENFORCEMENT
For an enforcement system to be successful, it must provide adequate incentive to pollutant sources to comply with emission and/or air quality regulations. With an ICS, establishment of incentive is especially critical because of the relative degree of independence the source has through its authority (albeit limited authority) to vary emissions. Adequate incentive exists if the regulation associated with the ICS: (1) Provides for adequate control agency surveillance of the source and/or its impact on air quality, (2) contains provisions enabling the control agency to establish legal liability if air quality standards and/or ICS emission regulations are violated, and (3) prescribes sufficient penalties to deter the source from allowing such violations to occur.
If the above conditions for adequate incentive hold, there are four approaches to enforcement that a control agency could pursue:
1. Enforcement on an air quality basis. The source operates the ICS and is held directly responsible for maintaining air quality standards in the vicinity of his facility.
2. Enforcement on an emission basis. The source operates the ICS and is required to vary emissions in accordance with emission "curtailment criteria." (Curtailment criteria are specific meteorological conditions or specific air quality levels at which the source must curtail emissions by predetermined amounts.)
3. A combination of approaches (1) and (2). The source is held directly responsible for operating in accordance with curtailment criteria, and simultaneously with assuring air quality standards are protected.
4. Control agency operation of the ICS. The agency, on an operational basis, determines when and in what manner the source varies emissions to attain and maintain air quality standards.
Enforcement on an Air Quality Basis:
This approach allows the source the greatest degree of independence and flexibility because the source operates the ICS itself, and its daily operations are not necessarily subject to control agency surveillance. The only stipulation is that the source must assume full responsibility for maintenance of the air quality standards.
Of course, as with enforcement on an emission basis, the curtailment criteria used by the source in its daily emission control decisions must meet prior approval by the control agency. In addition, the criteria are subject to periodic re-evaluation by the agency on the basis of how well the system is performing.
To assure that the standards are being met, real-time air quality data must be transmitted from the monitoring sites directly to the control agency, as well as to the source. The agency must have free access to all monitors to assure that they are properly calibrated and maintained. Enforcement actions are initiated if and when air quality standards (regulations in this case) are not met.
The number of sources desiring to participate in an ICS will be a major factor in determining whether air quality is the preferred basis for enforcement. In the case of a single isolated point source, the problems involved with air quality as the basis for enforcement (e.g., establishment of liability if air quality standards are exceeded) are minimized. Given a single isolated source, air quality monitoring would seem to offer a more direct approach to surveillance and enforcement than emission monitoring and engineering inspections because attainment of specific air quality levels (standards) is the principal objective.
The success of a system that bases enforcement primarily on measured air quality depends upon whether or not it can be shown, through analysis of the data, that a given source contributed a specific amount to the ambient concentration at a specific monitoring site. In a multiple-source situation, such a determination is difficult or impossible unless extensive emission data from the sources involved and appropriate detailed meteorological data from the area are continuously available. When more than one source is involved, it would probably be necessary, from an enforcement standpoint, to prearrange a legally binding distribution of liability.
With air quality as the principal basis for enforcement it is especially important that the number of air quality monitors be sufficient to provide a reasonable estimate of maximum ground-level pollutant concentrations due to the source in question. The eventual determination of what is "reasonable" will depend on (1) the cost involved with each additional air quality monitor, and (2) the acceptable degree of error in estimating concentration maxima. An acceptable tradeoff point between those two factors would have to be found. At any rate the required number (a dozen or more) of air quality monitors about each source would be much greater than is currently generally required.
The required number of air quality monitors might be considerably reduced if (1) meteorological dispersion models can be used in conjunction with air quality monitoring or (2) mobile sensors are employed, (3) or both. However, until validated dispersion models for the vicinity of the source in question are developed, a full complement of fixed and mobile air quality monitors would be required.
To be sure, the determination of optimum locations for placement of air quality monitors is difficult. The locations of maximum ground-level concentrations depend upon source characteristics, topography, meteorological conditions, and travel times before emissions reach ground level. The optimum network may be achieved only after considerable experience and adjustment of the locations of air quality sensors. However, it is at least as difficult to determine with any confidence (1) the degree of constant control (CCS) that would be required and (2) to demonstrate that air quality standards have been achieved as a result of that control.
Enforcement on an air quality basis, would, however, bring about a problem that does not exist with enforcement of emission regulations. Under emission regulations, the source is not necessarily subject to enforcement action if an air quality standard is exceeded in its vicinity, as long as it complied with the (fixed) emission regulations.
With air quality as the basis for enforcement the source must be held directly liable for violations of the standards if the source is to have adequate incentive to operate the ICS with the degree of diligence necessary to protect air quality. The problem is that for any large emitter, violations of short-term standards can easily occur due to the vagaries of the weather and the dependence of the success of an ICS on the skill and conscientiousness of the operators of the system.
The crux of the problem is how to legally enforce against such violations. Since a large emitter using ICS will likely cause air quality standards to be exceeded, ample incentives are needed to assure that the source would do its best to minimize the risk of such violations. Perhaps a graduated penalty system would be in order, penalties being assessed in proportion to the frequency and severity of violations.
Enforcement on an Emission Basis:
This approach to enforcement is similar to enforcement on an air quality basis in that the source operates the ICS; i.e., the source operators determine when curtailment criteria are met, and emissions are varied accordingly. However, in this ease, the control agency oversees the daily source operations. Source emission data, meteorological information, and air quality data must be available to the control agency (not necessarily on a real-time basis) so that it can determine if emissions are, in fact, being curtailed when the curtailment criteria indicate the need to do so.
Enforcement actions are initiated if the source does not properly (and promptly) respond to the curtailment criteria. Through this approach to enforcement it may not be possible to hold the source legally responsible for violations of air quality standards, as long as the source curtails emissions as dictated by the curtailment criteria. Such a possibility exists because the source will have been utilizing curtailment criteria approved by the control agency prior to initial acceptance of the ICS.
At any rate, enforcement entails a considerably different approach if emission regulations (for variable emissions) are involved, rather than air quality regulations. With an ICS, surveillance of source emissions and establishment of liability for violations would be a relatively arduous task because of the variations in emission rates that are effected at the source in response to continually changing meteorological conditions. Enforcement would be based at least in part on whether the source properly curtailed emissions as required by the regulations. The control agency administrative burden would be relatively complex because emissions would have to be continually matched with measured air quality, predicted air quality, and/or meteorological information to determine if the emission regulations are being complied with and whether the air pollution model is operating appropriately.
Enforcement on the Basis of Air Quality and Emissions:
This approach employs surveillance of both air quality and emissions as a basis of the enforcement procedure. The air pollution control agency monitors air quality to assure that the standards are being met. It monitors emission rates to assure that air quality standards are not jeopardized in areas where no air quality sensors are located and to afford some protection from unquantified adverse effects of the pollutants (see Tab. 4).
Control Agency Operation of the ICS:
Through this approach, the control agency operates the ICS itself, and dictates to the source when and in what manner to vary emissions. Through such an approach, the source is relieved of all direct responsibility for air quality, as long as it follows the instructions from the agency.
Enforcement actions are initiated if and when the source does not follow those instructions. Control agency personnel would be required to have extensive training, not only in meteorology and dispersion modeling, but also with regard to the source operations.
This approach allows the source no flexibility. The source is continually subject to direct operation orders from the control agency. Agency personnel would constantly oversee and dictate the source's control activities and possibly, depending on the type of operation, its production rate as well. This approach is highly paternalistic. It is so philosophically divergent from normal economic and industrial practices that it is unlikely to be acceptable to any source.
DISCUSSION
A situation where only two or three sources are involved needs consideration. As with the
case involving many sources, more complex emission regulations or some fixed diversion of liability would be in order to determine the proportion of the ground-level concentration at any given point is due to each source. The feasibility of enforcing the complicated regulations for such a system would have to be determined on a case by case basis.
In addition to all of the above considerations, there is the prominent fact that EPA and most state and local control agencies are ill-equipped to meet the enforcement demands placed on them by numerous ICS operations. If the use of ICS is not carefully restricted, the enforcement burden could easily become unmanageable. In addition, EPA will have to establish legal enforcement procedures and assign additional field personnel to intervene in those ICS situations where state and local control authorities do not act effectively to ensure adequate protection of air quality.
Any of the four approaches to enforcement, due to the relative complexity and inherent uncertainties of an ICS would impose a considerable administrative surveillance and enforcement burden upon the control agency. The burden is compounded if the source is located in rugged terrain, if more than one source participates in the system, or if the source is not isolated from other sources.
Terrain poses an extremely difficult problem because of the variety of subtle changes in dispersion characteristics found in such areas and the consequent effect on ground-level concentrations. Multi-source operations or operations in areas where a significant background of pollutant exists render the establishment of liability, the identity of the offender, or both highly uncertain.
In summary, enforcement considerations dictate that ICS be applied primarily, if not solely, to single facilities that are isolated from other sources of the pollutant involved or for multiple sources which agree beforehand to share responsibility for air quality violations on a fixed percentage basis. If there are measurable background concentrations of the pollutant, the source(s) desiring to use an ICS should assume responsibility for all, or at least a fixed percentage, of ground-level concentrations at the monitoring sites. From the standpoint of environmental protection (and enforcement, for that matter) the preferred approach to enforcement in any ICS is to base enforcement on air quality and emissions.
TAB 4. LEGAL POSITION OF ICS
The standard setting procedure for stationary sources of pollutant emission is covered by Sections 110, 111, and 112 of the Clean Air Act. Section 110 requires that each State submit a plan for the achievement and maintenance of NAAQS which "includes emission limitations and such other measures as may be necessary– ." Section 111 requires "the best system of emission reduction – taking into account the cost of achieving such reduction – " to be applied to new sources or source modifications "which contribute significantly to air pollution – ." Section 112 requires that "emission standards" be prescribed for sources of hazardous pollutants.
The only language in the Clean Air Act that would allow the use of dispersion techniques (ICS & tall stack) to achieve NAAQS are the words "and such other measures as may be necessary" in Section 110 (a) (2) (B), and the mention of "interim measures" and "available alternative operating procedures and interim control measures" in Section 110 (e) and (f) dealing with extension of the time for compliance with parts of the SIP.
There is no question that ICS qualifies as an "interim control measure" and as an "available alternative operating procedure" under 110 (c) & (f). EPA would be required by the Act to accept, and even to demand ICS to minimize pollutant concentration during SIP extension periods. Nor is there any question as to whether a reliable and enforceable ICS qualifies as "such other measures as may be necessary" in a case where no alternative CCS is available. Improved dispersion is the only alternative to emission reduction for lowering ground level pollutant concentrations.
The principal question involves the legality of rejecting dispersion control methods in favor of emission reduction when the former is sufficient to achieve NAAQS, preferred by the source, and allowed by the State.
EPA has rejected ICS in the past, but on grounds of reliability. An example of this approach is given in the preamble to the regulations proposed for non-ferrous smelters in Western States on July 27, 1972 (37 F.R. 16095). This policy is as follows:
At this time, it (intermittent control) is not considered an acceptable substitute for permanent control systems for attaining and maintaining national standards. Experience with systems employing intermittent process curtailment indicate that although air quality is unproved, violations of ambient air quality standards still occur. Additional experience with these systems may, however, in specific cases, improve their reliability.
Recent data from TVA (See Tab. 2) indicate that very significant reductions in violations of NAAQS can be achieved with ICS reductions such that no violations of the primary 24-hour SO2 standard occurred at the locations of 14 samplers. So the general rejection of ICS on reliability grounds is no longer supportable.
ICS could be rejected for some sources, confined to a supplementary role for others, and accepted only until more cost-effective CCS becomes available for still other sources if the CAA were interpreted to place value on emission reduction above and beyond NAAQS attainment. The CAA has not been so interpreted, to date. For example, the following exchange on this point took place during the oversight hearings before the House Subcommittee on Public Health and Environment, January 1972. The speakers are Subcommittee Chairman, Paul 0. Rogers, EPA Administrator, William D. Ruckelshaus, and Deputy Assistant Administrator, Dr. John T. Middleton:
"Mr. ROGERS. Let me suggest this is something we would like to go into. Section 110 requires inclusion of emission limitations. I notice on page 3, you do not mention that. Maybe it is covered, is it?
"Mr. RUCKELSHAUS. I am not sure I understand the question.
"Mr. ROGERS. In the States' implementation plan which they must submit as to how they will implement the law, the law requires the inclusion of emission limitations in that plan. I do not see you mentioning that. Perhaps you overlooked it. Is that a requirement in your guidelines?
"Mr. RUCKELSHAUS. The answer to your question, Mr. Chairman, is that we have told the State that while they may submit a strategy in which emission limitations are provided, if they can show that the standards can be met without emission limitations, then we will review the plan with that in mind.
"Now, the vast majority of the plans that have been and are being submitted, do contain and will require emission limitations.
"Mr. ROGERS. How will they know if they have to have emission limitations until something happens?
"Mr. RUCKELSHAUS. In virtually all of the plans that we have now, there are emission limitations.
"Mr. ROGERS. Why is that not a requirement since it was specifically stated in the law?
"It is my understanding this was changed at OMB, that you had it in your suggested guideline but it came back from OMB and it was not in it. Maybe I am mistaken.
"Mr. RUCKELSHAUS. It was simply amended to say that if a State could show they could meet the air quality standard without emission limitations, then there would not be such a requirement. Frankly, I do not know how they are going to do it, but if they can make such a showing–
"Mr. ROGERS. I do not know how they could do it either. It seems to me there was a requirement, and I hope it is still a requirement because it is required in the law.
"Mr. MIDDLETON. Mr. Ruckelshaus spoke to other opportunities to meet the standards. I think if you or your staff had the opportunity to look at the rules and regulations, which we will submit for the record, you will see under subpart A, 42.1 a description of what a control strategy is. I might just read that to clarify the point since it is one of concern to all of us. It says:
"'Control strategy means a combination of measures designated to achieve the aggregate reduction of emissions necessary for attainment and maintenance of a national standard including but not limited to measures such as emission limitations.' These are first on the list.
"Mr. ROGERS. And I would think the most important.
"Mr. MIDDLETON. That is why they appear first on the list.
"Mr. ROGERS. I cannot conceive of any plan coming in where they do not have some thoughts, some plan, some method of eliminating emissions, because this is where we are beginning to start one.
"Mr. RUCKELSHAUS. I have yet to see one which does not have any limitations.
"Mr. ROGERS. I would hope you would not approve any plan that does not include that within their proposal.
"Shouldn't each plan have that?
"Mr. MIDDLETON. We would not want to stop progress if there is an innovative idea.
"Mr. ROGERS. This does not stop any new ideas. It just says that if the new ideas do not work, we can stop the emissions.
"Mr. MIDDLETON. The Administrator will have to approve the plan.
"Mr. ROGERS. How are you going to effect anything or carry it out unless you know how to bring about a limitation of the emissions? Here we state it first in the law, and then it is left out of your guidelines.
"Mr. RUCKELSHAUS. I do not think it is left out.
"Mr. ROGERS. Then the intent is that they shall, is that correct?
"Mr. RUCKELSHAUS. It is certainly so stated in the regulations.
"Mr. ROGERS. If that is the clear understanding, fine, but I think that should be made clear, and I think the States should know that because certainly that was the intent of Congress and I am sure your intent in getting into it."
It should be clear from the remarks of Congressman Rogers that the intent of Congress in the Clean Air Act is that the State Implementation Plans must include emission limitations. It is also clear that the words "and such other measures as may be necessary" exclude the interpretation that emission reduction is the only acceptable means of meeting NAAQS. Between these boundaries to interpretation, there is a broad, unexplored territory.
This territory includes such questions as: Are emission limits required for each source or for the SIP generally? How much emission limitation is required before "other such measures" can be employed? On what basis is the proportion of ICS and CCS established? Is ICS an emission limitation control method?
There are two basic approaches to this legally unexplored territory. The first approach, which may be called the broad interpretation, views the CAA in its entirety and notes that emission limitation (Section 111) is required of new sources without regard to air quality unless NAAQS would be exceeded by a new source with the specified emission control. The criteria for control of new sources is that the "best system of emission reduction" be used provided that it has been demonstrated and that its cost has been taken into account. Emission standards are also required for hazardous pollutants (Section 112). Section 111(d) refers back to Section 110 as follows:
"(d) (1) The Administrator shall prescribe regulations which shall establish a procedure similar to that provided by Section 110 under which each State shall submit to the Administrator a plan which (A) establishes emission standards. For any existing source for any air pollutant (1) for which air quality criteria have not been issued or which is not included on a list published under Section 108(a) or 112(b) (1) (A) but (ii) to which a standard of performance under subsection (b) would apply if such existing source were a new source, and (B) provides for the implementation and enforcement of such emission standards. (Italic added.)
This section requires emission standards for existing sources of non-criteria pollutants established "by a procedure similar to that provided under Section 110." This clearly implies that emission standards are required under Section 110. This implication is made not only by the plain language of the Act, but also on equity grounds. Why should emission standards be required of existing sources of non-criteria pollutants when such emission standards are not required for criteria. pollutants? When Section 110(a) (2) (B) is read in this light, emission reduction is clearly the preferred control method, and "such other measures" are allowed only if emission reduction sufficient to meet NAAQS in the time specified (3 years) is unavailable or infeasible – or, in the words of the Act, only if they are "necessary."
The second interpretation, which may be called the narrow approach, focuses on the objective of Section 110(a) (2) (B) rather than the means of attaining that objective. The principal objective of an SIP is that it meet primary and secondary standards by the appropriate deadline. Several means have been suggested, including emission limitation land use controls and transportation controls, but Congress was careful to add "other such measures" and "but not limited to". Thus, any means may be employed provided the ends are attained. This interpretation is strongly suggested by Dr. Middleton in the testimony quoted above.
These, then are the legal arguments to be expected when this issue arrives in court. The environmental organizations will use the first argument coupled with the material presented in Tab. 4: Unquantified Effects of Pollutant Emission. Industry will use the second argument supported by cost-effectiveness arguments similar to those presented in Tab. 8.
One further legal point requires discussion. That is whether or not ICS is an emission limitation method. This is discussed in Tab. 1: Theory and Operation of ICS. No clear-cut answer can be given because emissions under ICS are reduced sometimes. If, however, ICS emissions are viewed over a period of months or years, emissions close to or even greater than before ICS are likely to be seen. Therefore, in a very practical sense, ICS is not an emission reduction control method, but an alteration of the timing of emissions so that ground-level air quality impact is reduced. ICS must be classified primarily, if not totally, as a technique for increasing pollutant dispersion different in operation but similar in effect to a tall stack.
TAB. 5. UNQUANTIFIED EFFECTS OF POLLUTANT EMISSIONS
There are three types of effects due to pollutant emissions that are not presently covered by NAAQS:
1. Known adverse effects which cannot be quantitatively linked to emissions,
2. Undesirable effects of emissions that are difficult to evaluate in terms of public welfare costs, and
3. Unknown cumulative effects of sustained or repeated exposure to low-level concentrations.
The first category covers the acid rain phenomenon, conversion of SO2 to suspended sulfates and the effect of suspended particulates on climate. Visible smoke, visibility decreases to below 5 miles, visible damage to vegetation, sundry other aesthetic effects, subtle ecological changes and effects on non-economic flora and fauna fill the second category. The third category obviously cannot be documented. It is related to the concept of damage threshold, basic to the setting of NAAQS. There may be no thresholds of adverse effect. Present NAAQS may only protect against the more obvious and rapid adverse effects.
Acid Rain
a. Evidence
Normal precipitation would tend to have a slightly acidic pH (pH 7 is considered neutral) due to CO2 absorption and subsequent carbonic acid formation. Because pH is a logarithmic function, a pH of 5 is 10 times more acid than a pH of 6 and 100 times more acidic than a pH of 7. In this light, normal carbonic acid rain has a pH of about 5.7. Data in Scandinavia show 200-fold acidity increases in rainwater in some parts of the country since 1956 with a low pH recorded at 2.8, a value some 2000 times more acid than normal precipitation.
A brief note on the atmospheric reactions of SO2 is needed. Where SO2 is not absorbed in plant or animal tissue, or brought to earth as particulate sulfate, it subsequently is oxidized to SO2 and ultimately reacts with atmospheric water to become sulfuric acid (H2SO,). The acid is then cleansed from the atmosphere by normal precipitation. The rate of the above reactions can vary from between a few hours (under conditions of high ozone, high humidity, and temperature, as well as high aerosol content) to four weeks under conditions antithetical to those cited above.
Studies in Scandinavia indicate that increases in acid precipitation levels at various points between 1955 and 1970 coincide with increased anthropogenic SO2 emissions.
Chemical studies on the nature of precipitation in the U.S. are far less complete than those in Scandinavia. However, where measurements have been made over the course of years, a trend similar to that noted in Scandinavia is noted.
Experiments in the Northeastern part of the U.S. include those at Hubbard Brook, N.H., the Finger Lake Region, N.Y. State, New Durham N.H., Hubbardston, Mass., and Thomaston, Conn.
All of those rural regions had weighted annual pH averages between 4.2 and 4.3. These data were found in the latter 1960's. New Haven, Conn. showed an average of 3.81 for the pH of rainwater, and Killingworth, Conn., 30 miles east of New Haven, had a pH of 4.3 during the same time period.
On the other hand, sections of the country where SO2 emissions are not present to any considerable degree, or where prevailing winds would tend to prevent air from polluted regions from arriving at some sites, pH values in rainwater are close to normal. This certainly indicates that as SO2 emissions increase, regions within a few hundred miles of the point source will show increased rainwater acidity. The National Center for Atmospheric Research (NCAR), an arm of NOAA, has performed an investigation of U.S. precipitation chemistry during the 1950's. Much of their data is unpublished, but available, and might be utilized to get a better handle on U.S. acid rain over a period of years.
b. Ecological effects
Because of the general chemical activity of the hydrogen ion, it can be considered as a non- threshold type species as defined earlier; i.e., its effects are cumulative and only the extent of damage depends on concentration. The ecosystem is threatened primarily in five ways by excess environmental acid.
(1) Direct Damage to Trees and Plants.
P. Gordon showed that pine needles inoculated with acid solution of pH<3.5 were dwarfed and that needles sprayed with acid solution <4.0 grew to one-half normal length. It was concluded that acid rain contributed to the tree dysfunction known as "short-long" conifer needle syndrome.
(2) Direct Damage to Microorganisms.
Most nitrogen fixing bacteria are primarily confined to alkaline sons. Therefore, a reduction in alkalinity could have an impact on the available soil nitrate used as plant nutrient. Furthermore, the breakdown of plant litter by decomposing bacteria varies as a function of pH. This implies that optimum decomposition and mineralization may be reduced due to soil pH changes.
(3) Indirect Damage to Biota.
Because calcium, magnesium, and potassium present in soil are essential plant nutrients, their loss, due to acid leaching, has serious environmental implications. This problem has been studied quite extensively by the Scandinavians over the past ten years or so.
The Swedes took soil samples from 200 locations and measured the pH, Mg, K, and Ca contents of the soil. Assuming a fixed acid deposition rate of 10 milli-equivalents of H+ ion/M2-year, and then computing the leaching from areas of different original cation concentrations, forest productivities are computed. An average of .4%/year is found, but is considered too high by a factor of two due to not considering cationic concentrations below the 5 cm top soil level.
Therefore, an average reduction of .2% per year with a worst case of .5% /year are calculated.
The extrapolation to 1972 shows an average of about .3 % /year due to an increase in SO2 emissions. These calculations exclude direct leaf damage from sulfate acid aerosols, and acid rain, as well as population changes among decomposing and nitrifying bacteria.
Furthermore, secondary effects could set in, e.g. seeds may not take in more acid conditions. Due to decreased growth of root systems, watersheds might deplete and soil begin to erode thus hastening the ecological deterioration of the forest.
(4) Effect on Arable Lands.
The acid problem on arable lands is mostly economic. The more acid deposited, the more alkaline fertilizers must be used, which create expenses both in terms of fertilizer needed and manpower needed to spread it. Also, it is not a certainty that all adverse effects could be mitigated by additional fertilizing, or that adverse side effects would not spring up because of the excess liming required.
(5) Effects on Lakes and Rivers.
In Sweden and Norway pH values of rivers have dropped as much as .5 pH units in five years (a seven-fold acid increase) with an average annual increase of .3 pH units. These data are for heavily polluted aquatic systems. In systems where the only pollutant source is precipitation, the average pH drop appears to be about .15 pH/5 years. Furthermore, because bicarbonate ions are being depleted (they act as buffers to mitigate the pH drop when acid is deposited), the rate of pH decrease is expected to begin a rapid increase shortly. The situation has become so bad that certain salmon species have ceased to breed.
Another study performed in Canada shows the urgency of the acid problem. The study deals with the acidification of the Lumsden Lake system 45 miles southwest of Sudbury, Ontario (site of much ore smelting activity). The lake pH dropped from 6.8 in 1961 to 4.4-5.8 in 1971 ... an 80-90 fold acidity increase! No manmade discharges empty into the lake, while organic bogs don't exist in the area, and pyrite concentrations are extremely low. Thus it was concluded that the pH drop was primarily due to SO2 fallout as sulfuric acid in rainwater.
The result of the pH lowering was the almost complete elimination of the following fish species from the lake: white sucker, lake trout, lake chub, trout-perch, slimy sculpin, burbot, and yellow perch. Some investigators documented terrestrial plant damage within a 5 mile radius of the smelters (including Linzon, Dreisinger, McGovern) as well as pond acidification within the 5 mile radius, but damage outside of a 15 mile radius was considered negligible. It is now apparent that the bicarbonate buffer system began degrading, while acidification rapidly advanced in the decade of the 1960's.
In the U.S. there have been few long term studies on inland bodies of water with respect to pH changes. One striking result is that reported by Schofield which shows that one large clearwater lake in the Adirondacks has gone from a 1938 pH of 6.6-7.2 with a calcium carbonate concentration of 12.5-20 mg 1 to a 1960 pH 3.9-5.8 with a carbonate concentration of 3.0 mg 1.
Other information on larger aquatic systems appears to show similar, though not nearly as drastic alkalinity reductions. Examples include a 7% alkalinity drop in Lake Michigan over the past 60 years, and sulfate increases in the Ohio and Illinois Rivers.
Suspended Sulfates
Recent preliminary results of the CHESS studies indicate adverse health effects due to suspended sulfates at 24-hour average concentrations in the range 8-12 ug/m2 (memo from Dr. Finklea to Assistant Administrator, Research and Monitoring, dated 1-12-73 Table 1). Figure 4-1 shows 1967 NASN suspended sulfate data for 160 urban and 31 non-urban sites. The two heavy lines indicate the combustion of geometric mean and standard geometric deviation expected to result in a 24-hour concentration of 8 and 12 ug/m2 on the second worst day during one year. As can be seen from the figure, the majority of both urban and non-urban sites exhibit sulfate concentrations above both lines. It is clear that significant adverse effects due to sulfates existed in 1967.
The genesis of suspended sulfates is not precisely known. It is certain, however, that a substantial part of observed sulfate concentrations is due to SO2 emissions.
Threshold Standards
When a standard is set based upon a threshold concept, the implication is that the environmental element has a defense system to purge or assimilate the noxious substance before any chemical damage can occur. However, when the amount of foreign substance becomes great enough, it overwhelms the purging apparatus. On the other hand, a non-threshold standard implies cumulative, non-reversible damage. Usually, the amount of substance plays a role in determining the rate at which damage is done; e.g., silting of a river bottom is harmful, but the damage done is worse if 20 tons/year are deposited than if 10 tons are deposited. This same mechanism holds true for most chemical reactions.
Threshold standards are based upon non-accumulation of pollutant in the environment and an immediate cause and effect when thresholds are exceeded. On the other hand, non-threshold standards are predicated on the fact that: (a) either damage occurs at any level; the lower the level the slower the damage occurrence; or (b) the pollutant accumulates in the biosphere, water, or soil, and then displays a threshold effect, in one of those media.
The important point to note is that threshold standards protect against high concentration in the atmosphere, while non-threshold standards protect against total quantity of atmospheric fallout.
Furthermore, the latter protection may afford biota indirect protection through conserving the medium within which the fauna or flora grows (e.g., water for fishes, soil for trees).
Air quality standards are intrinsically threshold type standards. Emission standards and limitations are non-threshold in that they limit or reduce all environmental impacts of the pollutant in question regardless of the degree of intensity at which those impacts are taking place.
Inadvertent Climate Modification
Based on existing evidence, it cannot be conclusively determined whether man is inadvertently modifying global climate to any significant degree through pollutant emissions into the atmosphere. Nevertheless, certain pollutants are suspected of causing such effects.
The weight of the evidence seems to indicate that global cooling at the earth's surface will result as atmospheric concentrations of suspended particulate increases. It is not clear, however, whether atmospheric particulate levels are even significantly increasing. Nor is it known to what extent controllable emissions contribute to the problem in comparison with agricultural dust, volcanic activity, and forest fires.
Anthropogenic injection of sulfur into the atmosphere, primarily in the form of SO2, is significantly compared to natural sources. Long-term or large-scale effects of SO2 are largely unquantified, largely due to natural conversion of atmospheric SO2 to other forms, such as suspended sulfate particulates. It is possible that some of the "measured" increases in background concentrations of suspended particulates are due, in part, to increasing sulfate concentrations. A major climatic effect of suspended sulfates is to decrease atmospheric visibility.
Visibility
The present particulate secondary 24-hour standard is based on preserving a 5-mile visibility. A view of the snow capped mountains would be seriously effected by such a visibility limit. The aesthetic component of particulate limitation is a major portion of the "non-degradation" issue now pending at the Supreme Court. The most likely cause of citizen complaints about air pollution is the sight of particulate emission (smoke). Although aesthetic effects may be difficult to price, they are nevertheless important to the public.
Ecological Impact
In the section on acid rain it is mentioned that breeding habits of some fauna are affected by environmental changes due to pollutants washed out of the atmosphere. Ecological change occurs fundamentally by differential reproduction rates. Any effect on the balance between the reproduction rates of the several animal and plant species in an ecosystem will change the entire system given sufficient time. The "welfare" effect of the decline of a non-economic species may appear negligible, but such decline is an indicator of environmental impact which probably portends much more serious systematic effects if neglected.
REFERENCES
[Footnotes Omitted]
APPENDIX TO TAB. 5. EXAMPLE OF AN INTERMITTENT CONTROL SYSTEM
Since September 1969, TVA has been exercising intermittent control at its Paradise Steam Plant. The program requires reduction in generating loads to reduce SO2 emissions during adverse meteorological conditions when a prescribed (threshold) ambient ground-level concentration would otherwise be exceeded.
Through several months of field investigation it was determined that the only meteorological condition likely to cause ground-level SO2 concentrations in excess of the threshold level was a "limited mixing" situation. Because only one "problem" situation prevails at Paradise, it is possible to apply a single model, based on 9 meteorological criteria, to determine when the emission rate must be reduced. The plant curtails emissions by reducing plant load. The criteria are:
1. Potential temperature gradient between stack top (183 m) and 900 m is > 0.46º C/ 100 m.
2. Potential temperature gradient between stack top (183 m) and 1500 m is > 0.51º C/ 100 m.
3. Difference between daily maximum and minimum surface temperature > 6º C.
4. Maximum daily surface temperature <33- C.
5. Maximum mixing height < 200 m.
6. Maximum mixing height > calculated plume centerline height.
7. Time for mixing depth to develop from plume centerline height to critical (maximum) mixing depth > 1.1 hours.
8. Mean wind speed between stack top (183 m) and 900 m is between 2.5 and 8.0 m/sec.
9. Cloud cover < 80 percent.
To determine when the criteria are met, meteorological measurements are made on a daily basis, utilizing a meteorological instrumented tower and aircraft, specialized support from the National Weather Service, and an ADP facility. The computer facility is used to determine when the criteria are met and, utilizing a TVA dispersion model, the necessary load reduction to assure that ground-level SO2 concentrations will not exceed the threshold level.
There are two occasions in which attainment of all prerequisite criteria would not result in the required curtailment: (1) when further curtailment would lead to system instability (e.g., blackouts) or (2) when the supply of firm power to customers would be interrupted. Neither of these situations has yet occurred.
Air quality data were used to determine the effectiveness of the ICS. A network of 14 SO2 monitors was established in a 22½° sector downwind of the plant, bounded by the 22½° and 45º azimuths. Before institution of the ICS (1/1/68 – 9/19/69), there were 10 violations of the secondary 3-hour NAAQS for SO2 and 8 violations of the primary 24-hour NAAQS. For a period of similar duration immediately following institution of the ICS (9/19/69 – 6/25/71), there were only 2 violations of the 3-hour standard and no violations of the 24-hour standard. Thus, the ICS was quite effective in reducing the number of violations at the TVA monitors.
It should be noted that the Paradise ICS utilizes but one (critical meteorological criteria) of several theoretical bases for emission curtailment; i.e., predicted and measured air quality may also be used, either in lieu of or in combination with meteorological criteria.
TAB. 6 COST EFFECTIVENESS
The principal arguments in favor of ICS, and the reason that smelter and power plant operators want to use ICS, is that permanent control technology is not available in every case to meet NAAQS and that ICS may significantly reduce the costs of meeting the air quality standards.
There is no doubt that if source operators were allowed to choose the method of achieving NAAQS, that many would choose a combination of tall stack and ICS, or would at least include these tactics in their control strategy. Kennecott Copper Corp. has estimated that the least cost control system for their smelters to attain the NAAQS is a combination of 40-80% CCS (using acid plants) plus ICS. To attain the standards by CCS alone would increase their costs by 50-350%. TVA has estimated that the cost of ICS is about 1/50 the cost Of CCS to achieve short-term S02 standards. (TVA power plants emit from tall stacks.) (See enclosures.)
There is no question of the cost-effectiveness of ICS when attainment of AAQS is the measure of effectiveness and the cost to the operator is the measure of cost. However, each of these calculations may be questioned.
ICS cost
The cost of ICS may be divided into three parts: Direct installation and operating costs to the source, lost production for the source and lost wages for its employees, and increased surveillance and enforcement costs to the public sector. The direct costs to the source include equipment for monitoring air quality and emissions and for varying the emission rate, modeling services, additional personnel, clean fuel (if that is the method of reducing emissions) and license fees.
TVA has estimated the cost of ICS operation at one plant at $282,000 initially and $103,000 (1970 dollars) annually, excluding the costs of clean fuel or load transfer. This estimate includes only 6 monitors and no telemetry equipment. If 20 monitors were used with telemetry, the cost would increase to about $500,000 initially and $150,000 annually. These cost increases include additional personnel as well as additional equipment.
Lost production may or may not be a serious cost factor depending on whether the method of emission reduction is production curtailment, on whether the plant operates at full capacity, on whether lost output can be recovered without serious cost penalties, and on the lead time for curtailment. If plant operation must be partially or totally curtailed at short notice during a period of high demand for production, this cost may be high indeed. On the other hand, if emission reduction is affected by switching to cleaner fuel, shifting the load elsewhere (in the case of power plants), or rescheduling periodic suspensions of operations for maintenance, then the lost production cost may be relatively small.
For example, Kennecott estimates that the least cost control system for their smelters should include about 60% CCS via acid plants with additional control by ICS. The CCS is required to reduce smelter operation curtailment time. If more ICS and less CCS were used, the smelter could become a serious bottleneck in the mine-to-market production system. Custom smelters, such as those operated by ASARCO, do not have mine operation overhead expenses to deal with, so their least cost operation may include more ICS and less CCS.
The cost of load switching to power plants may be very low if the number and duration of load reductions is low. For TVA's Paradise steam power plant, the estimated reduction frequency is 30 days per year with an average duration of 4 hours. If all the power needed during these curtailment periods is purchased from neighboring power plants not in the TVA system (the worst case), the incremental control cost due to the purchase of power would be about 1/4 of 1 percent of the cost of power production. This analysis assumes that the capacity to supply the additional power exists. If additional capacity does not exist, then either additional capacity must be built or emission reduction must be accomplished using a dual fuel system. Either of these alternatives would be considerably more costly than load switching.
The increased cost of surveillance and enforcement incurred by the public air quality control agency due to ICS may be considerable. The reliability of ICS in avoiding NAAQS violation is particularly dependent on vigorous policing (see Tabs 2 and 3). This effort will require additional resources at the control agency. The additional cost of policing an ICS may run as high as $130,000 the first year and $50,000 per year thereafter. These figures include 4-6 independent fixed samplers, a mobile sampler, telemetry equipment and 2-3 men full time to operate the equipment, check source operation and equipment calibration, and review source curtailment procedure. It is reasonable to expect the source to defray at least part of this public expense. This cost, then, should appear in the license fee paid by the source in connection with the control strategy approval procedure.
Indicators of effectiveness
The indicator of effectiveness used in virtually all cost-effectiveness calculations done by prospective ICS users is attainment of AAQS. The discussion of unquantified effects of pollutant emissions under Tab. 4 raises some doubt as to the adequacy of this indicator as a measure of environmental effectiveness.
Unfortunately, there is no good indicator of total environmental impact of emissions. The indicators currently used – annual emission, emission rate, and ambient concentration – each leave something to be desired. Measures of emission neglect the effect of atmospheric dispersion while measures of concentration usually are concerned with peak concentration or concentration at one point and neglect the distribution of concentration in time and space.
In order to quantitatively compare the effectiveness of ICS and CCS, or any control strategy elements for that matter, a superior indicator of environmental impact is desirable. An attempt is made in the Appendix to develop such an indicator. The result is an estimate of environmental impact based on emission rate, emission height, wind speed, stability and ceiling height.
Although this indicator is probably superior to either average emissions or maximum pollutant concentration, it is far from ideal because it assumes that environmental damage is linearly related to ground-level pollutant concentration and independent of what is on the ground, assumptions certain to be overly simplistic.
In summary, then, there is no available indicator of environmental impact that quantitatively includes all adverse effects of pollutant emissions. The minimum acceptable effectiveness of a control system is the attainment of NAAQS; this is the stated criteria of Section 110 of the CAA.
Total annual emission is an indicator that includes all environmental effects but is not quantitatively related to the expected environmental damage. Total annual ground-level concentration (see Appendix) is a potentially superior indicator, but it is complex and based on a simplistic damage function.
Cost effectiveness
In order to compare two or more control strategies on the basis of cost-effectiveness the measure of cost and the indicator of effectiveness must be mathematically related to produce one number or figure of merit for each control system. If attainment of AAQS is taken as a sufficient measure of effectiveness, then the cost of achieving AAQS is the only available indicator of cost- effectiveness, and the system with the lowest cost is the most cost-effective. The assumption underlying this procedure is that all control systems that meet AAQS are equally effective in preventing environmental damage.
If attainment of AAQS is taken as a necessary but insufficient indicator of effectiveness, and annual emission is the indicator of environmental impact, then the control system cost- effectiveness is given by (system cost) x (system annual emission after control). As in the previous example, the system with the lowest product is the most cost-effective. The assumption underlying this procedure is that if NAAQS are met, damaging impact of emission on the environment still occurs and that impact is directly proportional to annual emission. These assumptions are probably superior to the one necessary for the previous calculation, but they neglect the effects of dispersion on concentrations and of concentration on damage – serious omissions.
Total concentration (Appendix) may be substituted for annual emission in the cost-effectiveness calculation and thereby internalize the relationship between dispersion and concentration. The effect of this substitution would be to improve the cost-effectiveness calculated for ICS, tall stacks, or other methods that rely principally on improved pollutant dispersion. The assumption required for this procedure is that environmental damage is directly proportional to the time and intensity of ground-level pollutant concentration (dosage). This assumption is conservative in that there may be no damage due to very low or infrequent dosages.
To illustrate the use of these measures of cost-effectiveness, the following data have been assembled from various TVA estimates:
Estimated cost of scrubber (Experimental Widows Creek Plant) $42,000,000
Removal efficiency of scrubber 80%
Initial cost of ICS (Initial cost of program including six monitoring stations) $262,000
Operating cost of ICS (Excluding load switching cost) $103,000
These data must be manipulated somewhat before they can be made useful. The estimated scrubber cost is undoubtedly high due to its experimental nature. It will be reduced to $20 million. The annualized cost of scrubbers is about 25% of initial cost, so $5 million per year will be used as the cost of the scrubber.
Six monitors are insufficient for ICS air quality feedback. Increasing the number of monitors to 20 will increase the initial cost to about $500,000 and the operating cost to $150,000 per year. If the initial cost is capitalized at 10%, the annual cost of ICS (excluding load switching) is $200,000. The cost of load switching may be as much as $200,000. The total annualized cost of ICS to the plant will, therefore, be taken as $400,000.
Increased surveillance cost must now be added to the cost of ICS. Let us assume that it will cost half as much to police the system as to operate it and set the annualized agency cost at $75,000 per year.
The annual reduction of emissions due to the ICS system will be taken as zero. This assumption is based on the fact that load switching will only need to be performed a few days per year and that the load may increase on some other days due to ICS at other plants. With these very crude data the effects of alternative cost-effectiveness calculations may now be compared.
Case I – Cost Effectiveness as seen by the Operator: Measure of cost – annualized cost of control to the plant; Measure of effectiveness – attainment of NAAQS; CCS cost- effectiveness – $5,000,000; ICS cost-effectiveness – $400,000; Relative superiority of ICS – 12.5:1.
Case 11 – Cost-Effectiveness as seen by the Control Agency: Same measures of cost and effectiveness as in Case I except that $75,000 per year control agency cost is added to the ICS cost; Relative superiority of ICS – 10.5:1.
Case III-Cost-Effectiveness using Annual Emission as the Measure of Effectiveness: CCS cost-effectiveness ($5,000,000) (20%v)– 1,000,000 $-%; ICS cost-effectiveness ($475,000) (100%) – 475,000 $-%; Relative superiority of ICS – 2.1:1.
Case IV – Cost-Effectiveness using Total Concentration as a Measure of Effectiveness (Appendix).
Data are not available to calculate the indicator of effectiveness for ICS and CCS. The cost figures would be the same. The relative superiority of ICS would lie somewhere between Case II and Case III due to the fact that emissions from the ICS would be weighed less heavily than emissions from the CCS because they are released during periods favorable for excellent dispersion.
Summary
Cost-effectiveness calculations performed by plant operators, where attainment of AAQS is the measure of effectiveness used, will make ICS look attractive to large sources of SO2 and perhaps other pollutants as well. Extensive monitoring and real time feedback from the monitors, if required by ICS acceptance procedure, will increase ICS costs and reduce its attractiveness to some marginal sources. The internalization of increased agency surveillance costs in the form of license fees will also increase ICS costs, but probably not enough to make ICS unattractive to a large source.
While attainment of NAAQS is the minimum legal effectiveness required of a control system, it is a poor measure of control effectiveness, for all presently unquantified adverse effects of pollutant concentrations below NAAQS are neglected. An indicator of effectiveness that includes all environmental effects of emissions is annual emissions. When this indicator is used in cost- effectiveness calculations ICS and other dispersion techniques appear less attractive, although they still may be more attractive than CCS in some cases.
Annual emission is a very conservative indicator of environmental impact when it is coupled with the requirement that NAAQS must be met. If a control system employing ICS and/or tall stacks can meet NAAQS and is more cost-effective than a totally CCS system when annual emission is used as the measure of effectiveness, than acceptance of such a system is most probably in the public interest.
Total concentration (Appendix) is a less conservative and more accurate measure of environmental impact than annual emission. Its use would favor ICS and other dispersion techniques somewhat more than annual emission. It is questionable whether this increased accuracy is worth the greatly increased complexity of this indicator, especially as the accuracy attained is still far from perfect.