Muskie Congressional Record: Papermills

CONGRESSIONAL RECORD — SENATE

June 16, 1976

Page 18587

PAPERMILLS AND CLEAN AIR

Mr. MUSKIE. Mr. President, the Senate has heard much debate on the issue of protection of clean air areas through the significant deterioration provision in the pending amendments to the. Clean Mr Act.

Much controversy has been generated regarding the extent to which new industrial facilities can be built in clean areas, the possible sizes of such installations, and the cost of their construction.

The Environmental Protection Agency has just completed a report which provides answers to these questions for the paper industry. This analysis supports the fact that there is no basis for charges that S. 3219 would foreclose economic development in major portions of the country, or in any specific case, despite the grave concerns expressed by the kraft pulp and paper industry.

The EPA study is based on an analysis of new mills and expansions at sites selected as optimal by the industry in the absence of non-significant deterioration requirements.

Its major conclusions are:

House and Senate proposed amendments to prevent significant deterioration of air quality will not prevent the construction of new economically sized kraft pulp and paper mills. Rather, the Congressional proposals will require some new mills to use different air pollution control strategies such as further control of sulfur dioxide emissions, construction at an alternative site, use of taller stacks, etc.

The cost of non-significant deterioration proposals on the pulp and paper industry should be very small. Capital cost impacts in the 1980-1990 decade should not exceed 5% of industry baseline capital expenditures. Annual cost impacts should not exceed 4% for the capacity additions over the decade. The overall impact will be about a 1.5% increase in total annual costs for the industry nationwide in 1990.

All new kraft mills and capacity expansions announced since 1971, covering the period through 1978, could have been built as planned under both the Senate and House non-significant deterioration proposed air quality increments without additional control technology or altered fuel use. In fact, all capacity announced since 1971 could have been built under the Senate and House proposals even if all plants burned coal and met Federal new source performance standards (NSPS).

I ask unanimous consent that the entire study be printed in the RECORD following my remarks.

Mr. President, I recommend EPA's findings to my colleagues as yet another example of the complementary nature of prevention of significant deterioration and future industrial expansion. I reiterate that S. 3219 is designated not to halt development but rather to protect our air resources for the use of future generations.

There being no objection, the study was ordered to be printed in the RECORD, as follows:

PREFACE

The following is one of a series of technical reports prepared by EPA concerning the impact of alternative proposals to prevent significant deterioration of air quality on major industrial and energy air pollution sources. The purpose of these analyses is to assist Congress in its consideration of alternative approaches to achieving the goal of protecting and enhancing "the quality of the Nation's air resources so as to promote the public health and welfare and the productive capacity and cost impacts of its population." This report discusses EPA's analysis of the capacity of the Congressional proposals on the kraft pulp and paper industry.

The kraft pulp and paper industry has expressed concern over the impact of Senate and House proposed amendments to the Clean Air Act and of EPA's current regulatory approach to the prevention of significant deterioration (described in Appendix A) because of the location of kraft mills in fairly hilly terrain and the possible proximity of kraft mills to Class I areas. An analysis was made, therefore, using actual topography and location of a sample of mills selected as representative of sites meeting the industry's needs in the absence of a program to prevent significant deterioration of air quality. A determination was made of capacity affected and cost impacts under a number of control technology and fuel use assumptions.

EPA ANALYSIS OF THE IMPACT OF ALTERNATIVE NON-SIGNIFICANT DETERIORATION PROPOSALS ON THE KRAFT PULP AND PAPER INDUSTRY

Summary and conclusions

This report is an analysis of the impact of alternative non-significant deterioration proposals on future kraft industry expansion. Actual data on specific sites for new mills and capacity expansions at existing sites under investigation by the major pulp and paper manufacturers are closely guarded and have not been made public beyond 1978. In the absence of this information, historical and projected data through 1978 were used under the premise that non-significant deterioration proposals were in effect at the time the identified kraft pulp capacity was proposed for construction. Thus the mill sites identified reflect sites selected as optimal by the industry in the absence of non-significant deterioration requirements. New mills and capacity expansions from 1971-1978 were analyzed. For these mill sites, three capacity scenarios were examined:

Actual new mill and expansion capacities 1971-1978;

Duplication of total site capacity including existing and new announced expansions;

The construction of a prototype 1000 ton per day bleached kraft mill at each of the sample sites.

In addition, the cost impact of best available technological control requirements was assessed for projected capacity from 1980 to 1990.

1.1 General conclusions

The cost impact of non-significant deterioration proposals on the pulp and paper industry should be very small. Capital cost impacts in the 1980-1990 decade should not exceed 5% of industry baseline capital expenditures. Annual cost impacts should not exceed 4% for the capacity additions over the decade. The overall impact will be about a 1.5% increase in total annual costs for the industry nationwide in 1990.

1.2 Specific conclusions

Eighty-nine percent of the sample mill sites would have no difficulty in duplicating total site capacity even it the mills were forced to burn coal and just met NSPS. By examining total site capacity the analysis covered 43% of the industry's Kraft pulping capacity nationwide, including announced additions through 1978. In this manner the analysis had greater applicability. Capacity precluded at the affected sites would be minimal (less than 1%) if mills used or were required to control beyond NSPS (low sulfur oil were burned or stack gas scrubbers with locally available coal). The cost of compliance nationwide under both the Senate and House proposals totaled less than one percent of total capital and annual costs in the sample mills. The cost estimate is considered to be conservative and could be reduced if alternative sites or modified design parameters were introduced into the analysis.

Eighty-seven percent of the sample mill sites would have no difficulty in supporting a prototype 1,000 ton per day bleached Kraft mill burning coal meeting NSPS and generating all electricity onsite. Capacity precluded at the affected sites would be minimal (less than 1%) if mills were required to control beyond NSPS (low sulfur oil or stack gas scrubbers with locally available coal). The cost of compliance under both the Senate and House proposals nationwide totaled less than one percent of total capital and annual costs in the sample mills. This cost estimate is considered to be conservative and could be reduced if alternative sites or design parameters were introduced into the analysis.

If provisions requiring BACT were defined as requiring control beyond NSPS, the House and Senate proposals would increase the industry's capital requirements up to a maximum of $570 million in the 1980 to 1990 decade assuming all new additions burned coal and had to employ scrubbers. This is equivalent to a 5% increase in industry's capital needs. The comparable estimates for annual cost increases in 1990 were $173 million which is equivalent to an increase of 4% for the mills built between 1980 and 1990. If BACT were defined only to require compliance with current NSPS, as is possible under the Senate proposal, the cost impact would be significantly less for this period and is anticipated to be comparable to those impacts defined for the mills announced capacity additions in the 1971-78 period.

If mills are required only to meet existing NSPS, regional impacts would vary considerably in mills which would have to employ additional controls or alter fuel use in order to duplicate total site capacity and comply with proposed allowable air quality increments under the Senate and House proposals. Impacts identified in the analysis were concentrated in the northeast and mountain and Pacific regions ranging up to 10 and 16 percent of capacity potentially precluded in these regions under the Senate and 17 to 24 percent of capacity potentially precluded in these regions under the House Class I and II and if additional steps were not taken to reduce emissions.

Under the House proposal, the Class III increment would eliminate almost all of these regional capacity constraints. If BACT were defined as requiring control beyond NSPS, or if all of the affected mills used low sulfur oil or scrubbers with coal, almost all of the capacity noted above could have been built in these regions. These additional controls would not impose a significant cost (less than 4% of annual costs) and should allow almost any economically sized mill to be built. However, siting will have to be more selective in both the northeast and mountain and Pacific areas of the country. The other four economic regions had minimal to no impact attributable to the proposed Senate and House air quality increments.

2.0 Characteristics of sample kraft pulp and paper mills

The most ideal sites to analyze to determine the impact of proposals to prevent significant deterioration of air quality (NSD) are the ones currently under consideration by the major corporations in the pulp and paper industry. However, future site planning is a closely guarded secret and it has not been possible to gather this information from individual firms or the American Paper Institute (API). An alternative approach which is used in this report is to analyze historical data under the hypothetical premise that NSD regulations went into effect at the time the mills were proposed for construction. Locational decisions during this period actually ignored NSD considerations and it is, therefore, possible to isolate whatever impact NSD could have had in terms of capacity precluded and additional pollution control costs.

The kraft mill sites selected for analysis cover all known expansions and new mills since 1971 including announced capacity expansion to be completed by 1978. A sample of 46 kraft mills representing 40% of all kraft mill operations was identified in this manner. (See Table 21.) The sample indicates that most of the capacity growth took place through expansion at existing sites. New mills represented only 9 of the mill sites and 33% of sample capacity. (See Table 22.) Existing mills do not have the site flexibility of new mills but are generally much smaller additions. Expansions at existing sites averaged 215 tons per day compared to 633 tons per day for new mills.

[Tables omitted]

Capacity growth is expected to increase during the next 8-year period. Increases in total kraft pulping capacity of 32,390 tons per day from 1978 to 1985 as contrasted to 17,067 tons per day contained in the sample have been projected by Arthur D. Little, Inc. Therefore, the total sample capacity studied in this report was reviewed under a number of capacity scenarios to adequately reflect future growth requirements of the kraft industry over the next decade.

2.1 Alternative capacity scenarios analyzed

The limitations of using historical data to predict future activity was recognized. A different set of sites will assuredly be involved in future capacity expansions. It is not expected, however, that the pattern of future development will change markedly in the foreseeable future, i.e., the majority of additions will take place at existing sites and will be generally smaller in nature than entirely new mills. At some point, however, economies of scale or woodland constraints will limit this form of expansion for future development. Therefore, several alternative cases were explored.

As a second capacity scenario, the total capacity at each site was examined assuming that the total site capacity was new. This essentially had the effect of increasing the average site size from 371 tons per day to 974 tons per day. In addition to the nine new mills included in the sample, an examination of total site capacity at sample mills supporting expansions was also considered to be a realistic indicator of possible future kraft industry impacts as these capacities are indicative of capacities actually supportable by total woodland supplies at particular sites, although not indicative of a total national increase in capacity.

A third case was evaluated assuming that the capacity additions at each site were 1,000 tons per day. Although the average new mill size is 633 tons per day, some of the new mills under construction are as large as 1,000 tons per day. It is anticipated that new mills will be increasing in size. It was, therefore, determined that as a worst case the analysis should assess whether a prototype 1,000 ton per day bleached kraft mill could be built at the sample sites under alternative non-significant deterioration proposals. This analysis was made, however, ignoring capacity limitations imposed by woodland supply which in some cases would not support a 1000 ton per day mill.

2.2 Comparison of sample and U.S. kraft industry

Tables 24 and 5 compare the sample capacity to total U.S. kraft pulping capacity. As can be seen in Table 25, considering total capacity at sample sites, the average size mill at the 46 sites in the sample is close to the national average mill size, i.e., 888 tons per day nationally compared to 956 tons per day in the sample.

Using total sample site capacity, the analysis covered 32% of national kraft pulping capacity in 1970 and 43% of national capacitythrough 1978. A comparison of the regional distribution of sample site and national capacity totals as of 1970 indicates coverage ofat least 17% of the mills and at least 18% of the capacity in each region of the country.

2.3 Fuel use in sample and kraft industry

Most of the existing mills in the sample burn oil, gas or bark in the power boilers. Only 12% of the mills burn coal; coal and/or bark (6%); or burn coal, oil and bark (6%). This breakdown is comparable to the American Paper Institute's reported statistics for the industry as a whole. A comparison and regional breakdown are presented in Tables 26 and 27.

Fuel mix can make a significant difference in emissions from the mills. As can be seen in Table 28, a mill burning low sulfur oil or using a scrubber with coal emits only 40% of the SO emissions of a mill of comparable size burning coal just meeting New Source Performance Standards (NSPS) in the power boiler. Comparing the more typical case, a mill burning oil or oil and bark will emit only 70% of the SO emissions of a mill comparable size burning coal just meeting New Source Performance Standards. Therefore, assumed fuel mix will have an important impact on the results of the analysis. This report includes an analysis of the impact of alternative non-significant deterioration policies assuming that the mills use what they actually are burning, and also assesses the impact of the contingency that national energy policy will dictate the increased use of coal.

In addition, as indicated in Table 28, the purchase of electricity from offsite sources can also reduce emissions onsite. However, because almost all of the sample mills generate electricity onsite, the "purchase" option which is an alternative means of reducing emissions was not specifically analyzed.

3.0 Methodology

Using actual topography and location of the sample of 46 kraft mill sites, results from EPA approved air quality impact modeling were used to determine maximum allowable craft pulping capacity at each of the sites under alternative non-significant deterioration proposals.

Model results for two prototype bleached kraft mills in varied terrain and under three meteorological situations were used as a basis from which impact calculations were made. The methodology used to apply the air quality impact model results to the sample was devised solely as a screening procedure. Modeling would have to be performed for individual mills to determine if a restriction, in fact, exists. Nevertheless, the procedure was conservative and should give a reasonable approximation to the results that could be obtained through individual modeling at each of the mill sites.

Analyses were performed for each site using the three site capacity scenarios and varied fuel use assumptions described in Section 2.0. The different factors examined in the study are summarized in Table 31. A summary of the provisions of the House and Senate proposals is contained in Appendix A.3.1 Emission assumptions

Two prototype bleached kraft mills were modeled with 1,000 and 400 tons per day of capacity, respectively. Under all of the non-significant deterioration proposals, new kraft mills and modifications must use best available control technology (BACT). Any existing national new source performance standard that is applicable would meet this requirement under EPA's regulations. A more stringent definition might apply under the Senate and House proposals, principally with respect to the allowable emissions from the power boiler. In order to analyze the impact of the separate provisions under both the Senate and House non-significant deterioration proposals, the impact of proposed air quality increments under the area classification schemes are assessed assuming new additions need only comply with NSPS. Mills requiring additional control beyond NSPS in order to comply with the applicable increments are then identifiable. The economic analysis section assesses the costs to the industry of employing additional controls beyond NSPS.

Emission and plant design assumptions generally were based upon prototype kraft mill characteristics described in the report prepared by Environmental Research and Technology for the American Paper Institute (API) on the Impact of Non-significant Deterioration Proposals. The report was used principally to estimate fuel requirements and plant design characteristics. BACT for SO from the kraft recovery furnace was assumed to be 30 ppm (1.6 lb. per ton or 67 lb. per hour). Emissions for the coal and oil combustion cases were calculated directly from existing or tentative NSPS for kraft mills and steam generators, except that for particulate matter from oil combustion, the AP42 uncontrolled emission factor was used.Where bark is being burned, the API report supplied the necessary data, adjusted to ascount for the lower SO emissions from the recovery furnace and lower particulate emissions from oil combustion, as discussed above.

LESS THAN FULL USE OF THE CLASS II AND CLASS III INCREMENTS

Half the Senate Allowable Class II.

Collocation of Kraft Mills in Sample.

Collocation of Kraft Mills and Sample of '74 New Power Plants.

3.2 Air quality modeling

Four variations of air quality impact models were used to assess the potential air quality impact of prototype kraft mills depending upon the critical terrain and relevant meteorological conditions around the mill.

3.2.1 Class II and III impact analysis flat or moderate terrain case

EPA's Single Source Model (CRSTER) was run for both the 1,000 and 400 ton per day prototype mills with actual meterological data from sites in the northeast, northwest and southeast. The sites were chosen as representative of regions of the country where kraft pulping capacity is concentrated and where new mills have recently been constructed: Concord, New Hampshire (Northeast); Huntsville Alabama (Southeast); Portland, Oregon (Northwest).

The CRSTER model is used where terrain features are below the mill stack height. The results are summarized in Appendix B.

TERRAIN EXCEEDING STACK HEIGHT

Where terrain is above the stack height but below the height where plume impingement was assumed to occur, maximum concentrations were assumed to occur under high wind, neutral stability conditions (see Appendix B).

PLUME IMPINGEMENT CASE

The EPA Valley model was run for the prototype 1,000 and 400 ton per day kraft mills to determine 3 hour and 24 hour concentrations for SO. The Valley model is applicable where terrain is equal to or greater than the combined height of the stack and plume under stable conditions. Air pollution concentrations are highest under this case because the plume impinges directly on elevated terrain. Meteorological assumptions and model results are presented in Appendix B. It would be noted that the Valley model uses worst case assumptions as input rather than actual meteorological conditions.

3.2.2. Class I impact analysis

FLAT OR MODERATE TERRAIN

The Class I analysis is concerned with long distance transport. A Gaussian plume mode assuming limited mixing within inversion layer was used (see Appendix B for specific assumptions and results).

PLUME IMPINGEMENT CASE

As in the case of Class II and III, described in Section 3.2.1, the EPA Valley model was run for the prototype 1,000 and 400 ton per day kraft mills to determine 3 hour and 24 hour concentrations for SO2. The Valley model is applicable where terrain is equal to or greater than the combined height of the stack plume as described in the previous subsection.

3.3 Application of model results to sample mill sites

As indicated in the previous section, terrain makes an important difference to predicted air pollution concentrations from a given source. Each sample mill, therefore, was mapped to determine whether critical topographical features exist at the site. UTM coordinates, and where available, longitude and latitude in seconds, were used to obtain as precise location of the mills possible.

Class II and III increments were examined within a 7 km radius (defined by the minimal air quality concentrations anticipated beyond this distance by the various models discussed above and detailed in Appendix B) to screen bills with:

a. stacks exceeding terrain by 100, 300, or 500 feet. [i.e., the "flat or moderate terrain case" were terrain is at or below the top of the stack]

b. terrain between the top of the mill stack and height of the plume rise under stable conditions

c. terrain above the stock plus plume rise. [i.e., the plume impingement case].

The distance and altitudes of critical terrain features were recorded and model results applied.

For case (a) the CRSTER model results were applied to determine the concentration from a 1,000 ton per day mill meeting NSPS. In no case would there be a capacity limit imposed on mill sites with these terrain characteristics due to the low predicted concentrations. Therefore, such mills were categorized as not constrained by Class II and III increments. For case (b), the Turner model results were used to predict air pollution concentrations. Where there was a terrain feature that led to plume impingement (case c) at a further distance within the critical 7 km radius, the high wind, neutral stability results were compared to the Valley model results used for case (c) and the higher concentration was selected. For case (c), the Valley model results were applied. To determine maximum allowable capacity at a site under the terrain conditions (b) and (c) where potential size limitations may arise, the concentrations of the graphs in Appendix B for 100 lb./hour were adjusted for different fuel mix assumptions from Table 28 and any levels above allowable increments were scaled proportionately to meet the increment and reflect reduced size below 1,000 tons per day. Similar scaling factors were used to adjust for allowable capacities greater than 1,000 tons per day.

Maximum allowable capacities for the sites under the various allowable increments and fuel mixes were then compared to the three capacity scenarios to identify any capacity limitations or fuel use constraints. Sample mill sites also were mapped to determine the direction and distance of Class I area designations. If the Class I area was located at a distance within which the applicable Class I increment could be exceeded, the terrain was determined within the Class I area relative to the stack height of the sample mill. Depending upon whether significant terrain characteristics existed, the Valley model or Limited Mixing Model results were applied.

4.0 Class II and Class III impacts

4.1 New mills and capacity expansions announced since 1971

All announced new mills and capacity increases in the Kraft industry since 1971, which represented a total capacity of 17,067 tons per day, could have been built at planned sites under both the Senate and House Class II increments if the non-significant deterioration programs were applicable. No additional control technology would be required nor altered fuel use. All announced capacity increases could even have been built if all the new mill and additions burned coal meeting NSPS. The conclusion is unaffected if actual stack height, i.e., 200 feet instead of the assumed stack heights of 300 feet, is used in the analysis. (See Tables 41 and 42.)

4.2 Total capacity at sample mill sites

Under the Senate Class II increment, total capacity at sample sites, including announced additions since 1971 could be duplicated without restriction (using planned fuels with no additional control technology) at all but five mill sites. (See Table 41.) Three of the mills are in the northeast and two are in the mountain and pacific region. Capacity that would have been precluded it the mills did not alter fuel use would be 1,460 tons per day of capacity, only 3.2% of the total sample. This represented 632 tons per day in the northeast (10% of regional capacity in the sample), and 828 tons per day in the mountain and pacific region (16% of regional capacity in the sample). If the five mills burned low sulfur oil or used a stack gas scrubber with coal only 160 tons per day of capacity in the northeast at two mill sites would have been precluded representing only .04,percent of total sample capacity or 2.6 percent of northeast regional capacity in the sample. Under the most conservative and unlikely assumption that all of the sample mills burned coal just meeting NSPS, all but the five mills could duplicate total site capacity under the Senate Class II increment without any additional pollution control beyond that which is already required under the Clean Air Act (NSPS). To prevent a total of 2,902 tons per day of capacity representing 6.5% of the total sample capacity from being precluded, these five mills would have to employ a stack gas scrubber in order to burn locally available coal. With the use of a scrubber only 160 tons per day would be precluded.

Under the House Class II increment total capacity at sample sites, including announced additions since 1971 also could be duplicated without restriction (using planned fuels with no additional control technology) at all but five mill sites. (See Table 42.) Three of these sites are in the northeast and two are in the mountain and pacific region. Assuming actual fuel burned at these sites, capacity limitations at the five mill sites total 2,240 tons per day in the northeast (17% of regional capacity in the sample) , and 1,211 tons per day in the mountain and pacific region (24% of regional capacity in the sample). If the five mills that were affected used low sulfur oil or a scrubber with coal, the total sample capacity would still be limited at four mill sites. However, only 473 tons per day or 1% of the sample capacity nationwide would be precluded.

This represents 299 tons per day of capacity in the northeast (5% of regional capacity in the sample) , and 174 tons per day of capacity in the mountain and pacific region (3% of regional capacity in the sample) which would be precluded if industry capacity were duplicated under the non-significant deterioration requirements.

If the regions with the five mill sites were reclassified Class III under the House NSD proposal, capacity at only one mill site would have to use lower sulfur oil or stack gas scrubbing with coal to prevent 50 tons per day of capacity from being precluded.

Under the most conservative and unlikely assumption that all of the sample mills burned coal just meeting NSPS, all but five mills could duplicate total capacity under the House Class II increment. To prevent 3,121 tons per day or 7 percent of total sample capacity from being precluded if the mills burned coal, the five mills would have to employ a stack gas scrubber which would leave the capacity impacts noted above. However, if the regions with these mill sites were reclassified to Class III under the House proposal, capacity at only one mill site would have to use either lower sulfur oil or scrubbers with coal to prevent 50 tons per day of capacity from being precluded.

4.3 Prototype 1,000 ton per day bleached kraft mill

1,000 ton per day bleached kraft mills burning coal to generate all electrical capacity onsite could be located at all but five (under the Senate Class II increment) or six (under the House Class II increment) of the 46 sites in the sample. Three or four of these sites are located in the northeast and two are located in the mountain and pacific region. An increase or decrease in stack height would not change the results.

With the use of a scrubber or low sulfur oil at the mills, however, all but two mill sites could expand up to 1,000 tons per day under the Senate Class II increment. Both of the affected mill sites are located in the northeast and are located in extremely steep terrain (i.e., 6.4 and 11 percent slope, respectively). Capacity at these sites is limited to 698 and 888 tons per day, respectively. This represents a loss of 1 percent of the total sample capacity if it were assumed that a 1,000 ton per day mill were built at each of the 46 sites in the sample.

With the use of a scrubber or low sulfur oil, all but three of the mill sites with capacity constraints using coal meeting NSPS could expand up to 1,000 tons per day under the House Class II increment. Two of these mill sites are located in the northeast and one mill is located in the mountain and pacific region. Capacity at these sites is limited to 578, 688 and 828 tons per year, respectively. This represents a loss of 2% of total sample capacity if it were assumed that a 1000 ton per day mill were built at each of the 46 sites in the sample. Total site capacity of 1,000 tons per day could be built at all but two mill sites if the area around six mill sites were reclassified to Class III under the House NSD proposal. Capacity precluded at the two mill sites in the northeast would only be 432 tons per day all of the 1,000 ton per day capacity and could be built if low sulfur oil were used. Similarly, if coal were required as a fuel, and three of the mills were to use stack gas scrubbers, reclassification to Class III would allow all of the 1,000 ton per day capacity to be built.

5.4 Impact of allowing less than full use of the class II increment

5.1 Half the allowable class II increment

Some industry spokesmen have expressed concern that faced with an intractable air quality limit under which a State would have to absorb reasonable economic growth, a State might consider limiting a given industrial source to only half of the Class II increment. While none of the non-significant deterioration proposals limit the full use of allowable increments and there is no precedent for the imposition of such limitations in implementation measures already established under the existing Clean Air Act requirements, the analysis of this scenario was performed.

5.1.1 New mills and capacity expansions announced since 1971

All new mills and all but one capacity expansion in the sample could be built if the capacity additions were only permitted to utilize half of the Senate Class II increment. Ten tons per day that would otherwise be precluded could be built if low sulfur oil were used at the mill. All but three mill expansions could burn coal meeting NSPS and generate all electricity onsite and still comply with these requirements. It would be necessary to use a stack gas scrubber at the three mills if capacity additions were both forced to shift to coal and limited to half of the Class II increment as defined under the Senate proposal. All of these mills are in the northeast.

All new mills and all but two expansions could be built if the mills were only permitted to utilize half of the House Class II Increment. 102 tons per day at the two mill sites that would otherwise be precluded could be built if low sulfur oil were used. All but four mill expansions could burn coal meeting NSPS and generate all electricity onsite and still comply with these requirements. It would be necessary to use a stack gas scrubber at the four mills if capacity addition were both forced to shift to coal and limited to half of the Class II increment as defined under the House proposal. Three of these mills are in the northeast and one is in the north central region. If areas around all of the affected mill additions were reclassified to Class III, all of the mill expansion could burn coal meeting NSPS and comply with half of the allowable Class III increment.

5.1.2 Total capacity at sample sites

Total capacity at sample sites could be duplicated under the Senate Class II proposal at all but five mill sites if the mills were limited to half the allowable increments. Three of these sites are in the northeast, two are in the mountain and pacific region. Capacity that would be precluded under half the Senate Class II increment without the use of additional control technology or altered fuel use would total 3,166 tons per day or 7% of capacity nationwide. This would be comprised of 1,363 tons per day of northeast capacity (23% of regional capacity in the sample) and 1,803 tons per day of mountain and pacific capacity (35% of the regional sample). If low sulfur oil were used at five of the mills in the sample, constrained capacity would only be 1,594 tons per day under half the Senate Class II increments representing a loss of .5 percent of total sample capacity at the 46 sites. The regional breakdown of precluded capacity under half of the Senate Class II increment would be 242 tons per day in the northeast region (4% of regional capacity in the sample) and 1,352 tons per day in the mountain and Pacific region (26% of regional capacity in the sample). The same results would occur if all but five sample mills burned coal meeting NSPS and the five mills employed stack gas scrubbers.

Under half of the allowable House Class II increment, total sample capacity that would be precluded without the use of additional control technology or altered fuel use would total 3,794 at five sites or 8.5% of the total sample site capacity. The regional breakdown would be 1,740 tons per day of capacity in the northeast (29% of regional capacity in the sample) and 2,054 tons per day of capacity in the mountain and pacific region (40% of regional capacity in the sample).

If all five mills used low sulfur oil, 2,371 tons per day (53% of total sample capacity) would be precluded under half of the allowable House Class II increment. This represented 834 tons per day of capacity in the northeast (14% of regional capacity in the sample) and 1,537 tons per day in the mountain and Pacific region (30% of regional capacity in the sample) . Half of the Class

III increment under the House proposal would have the same impact as the full House Class II. If all sample mills were forced to burn coal, seven mill sites would require stack gas scrubbers in conjunction with locally available coal if only permitted to utilize half of the allowable House Class II increment. Once again, if allowed only half of the Class III increment under the House proposal, capacity limitations and the need for additional control technology would be the same as that noted for the full Class II increment.

5.1.3. Prototype 1000 ton per day bleached kraft mill

If only half of the Senate or House Class II increment were allowed and if a 1,000 tor per day bleached kraft mill burning coal meeting NSPS and generating all electrical capacity onsite were built at each of the 46 sample sites, only six mill sites would experience capacity constraints. Four of these sites are in the northeast, two are in the mountain and Pacific region.

The use of low sulfur oil or a scrubber would eliminate constraints at one of the northeast mill sites under half of both the Senate and House Class II increments and would eliminate constraints at one of the mountain and Pacific sites under half of the Senate Class II. Affected capacity at the other four or five mill sites under these assumptions would be reduced to 1,910 and 2,349 tons per day, respectively, under half of the Senate and House Class II increments. This represents a loss of 4.2 percent and 5.1 percent of total capacity, respectively, under half of the Senate and House Class II proposals if all 46 mill sites supported 1,000 tons per day of capacity. Impacts under half of the allowable House Class III increment, assuming affected mills employ control beyond NSPS are comparable to those impacts identified under the full use of the allowable House Class III. These impacts were significant, 50 tons per day at one mill site.

5.2 Collocation of kraft pulp mills

The modeling studies indicate that at least two 1,000 ton per day Kraft pulp and paper mills fueled by coal meeting NSPS could be located at one site in areas of flat or moderate terrain. In areas of hilly terrain, however, the collocation of two such mills would require some spacing or additional control technology. Two 1,000 ton per day mills meeting NSPS would have to be located 20 to 22 km. apart under the Senate and House Class II increments, respectively, in hilly terrain. The House Class III would facilitate collocation in hilly terrain and reduce this distance to as close as 13 km. Mills could employ additional control technology beyond NSPS to permit siting at closer distances. For example, two such 1,000 ton per day mills controlled beyond NSPS would have to be located only 13 to 16 km. apart in hilly terrain in the East and 5 to 7 km. apart in the West under the Senate and House Class II increments. Again, the House Class III increment in conjunction with control beyond NSPS would allow siting of the 1,000 ton per day mills 3 to 10 km. apart in hilly terrain in the West and East.

An analysis of the distances between the 46 industry sites under study indicated that only one mill site in the northeast would have been limited by the location of a second mill within close proximity to it. However, the limitations imposed on the mill capacity were identical to those imposed by the full Class II increments due to the topography around the mill. Only one of the two mills were affected because the topographical features imposed limitations for winds blowing in only one direction.

The pulp mills tended to be located in remote areas with ample spacing between mills to avoid potential problems.

5.3 Collocation of kraft pulp mills and power plants

None of the mill sites in the sample would have been limited because of proximity to coal-fired power plants.

6.0 Impact of class I area designations

6.1 New mills and capacity additions announced since 1971

No announced expansion or new mill built between 1971-78 would have been affected by the Senate or House Class I mandatory or discretionary area designations. The results are summarized in Table 41 and 42.

6.2 Total capacity at sample mill sites

6.2.1 Senate mandatory class I areas

Total capacity at sample mill sites could be built without altered fuel use or control technology given the current Senate mandatory Class I area designations. All but two mills would be unaffected even if total capacity at the sample sites including announced additions through 1978 burned coal meeting NSPS with complete onsite electrical generation. One of these sites is in the northeast, the other is in the southeast. The total capacity of the affected mills could be built, however, if they burned oil and bark. Therefore, even it all mills were forced to burn coal, the one mill in the northeast and one mill in the southeast, with total potential capacity losses representing only 2 percent of the sample capacity which could be precluded if the mills just met NSPS, could be built if scrubbers were used with the coal.

6.2.2 House mandatory and discretionary class I areas

Total capacity at sample mill sites could be built given the current House mandatory and discretionary Class I designations. Only one mill in the sample in the west south central region would have to take any additional steps to comply with non-significant deterioration requirements by burning low sulfur oil or using a scrubber with coal. If forced to burn coal, three mills would have been affected by the House Discretionary Class I area designations if total capacity at the sites including announced additions through 1978 burned coal meeting NSPS with complete onsite electrical generation. Two of these mills were in the northeast, one in the west south central area. The total capacity of the two affected mills in the northeast could be built if oil were burned or bark were burned along with coal. All of the 1,400 ton per day mill capacity in the west south central could be built if low sulfur oil were used or if a stack gas scrubber were used with the coal.

6.3 Prototype 1,000 ton per day bleached Kraft mill

6.3.1 Senate mandatory class I areas

One mill site would have been affected by Senate Mandatory Class I designations if a 1,000 ton per day capacity bleached kraft mill burning coal meeting NSPS and generating all electricity onsite were built at each of the sample sites. The capacity at the site would be limited to 541 tons per day. A 1,000 ton per day mill could be located even at the affected site if the mill burned low sulfur oil or used a stack gas scrubber with the coal.

6.3.2 House mandatory and discretionary class I areas.

Prototype 1,000 ton per day bleached Kraft mills and could be built at all sample sites if at three affected mill sites the mills used either low sulfur oil or a stack gas scrubber in conjunction with locally available coal. Two of these sites were in the northeast, one was in the west south central regional. Under the most conservative assumptions, capacity at three sites could be limited to 655, 655 and 721 tons per day, respectively, if the prototype 1,000 ton per day bleached Kraft mills used coal meeting NSPS.

7.0 Cumulative capacity and economics impacts of alternative non-significant deterioration proposals.

Cumulative impacts on the Kraft industry of alternative non-significant deterioration proposals are summarized in Tables 41 and 42. As can be seen in the tables once additional pollution control is employed at those select mills whose capacities are adversely affected by the allowable air quality increments, capacity constraints are minimal, less than one percent of sample capacity. The important issue that is explored in the following analysis is the cost of meeting the applicable air quality increments and best available control technology provisions under non-significant deterioration proposals and any regional effects that would lead to regional advantage dislocation or particular

7.1 Costs of compliance.

To determine the price and cost effects of compliancewith non-significant deterioration proposals, baseline costs were compared to the two pollution control scenarios: low sulfur oil and the use of a stack gas scrubber in conjunction with locally available coal. To simplify the analysis it was assumed that the entire capacity of the mill would be fueled by low sulfur oil or supported by a stack gas scrubber if it were determined that any capacity would be constrained by the applicable air quality increments. This means that the costs reported for meeting the proposed air quality increments are conservative as only partial use of the low sulfur fuel or stack gas scrubber is necessary in most cases.

Baseline industry costs are estimated using a new 800 ton per day bleached Kraft market pulp mill. In mid-1975 dollars, the capital costs were assumed to be $196 million and operating and maintenance costs were estimated at $181 per ton per day. Annual costs were calculated to reflect both operating and maintenance costs and annualized cost of capital at 18%. These numbers were derived by Arthur D. Little Inc. The assumed new mill costs were used solely for two of the capacity scenarios, i.e., total sample site capacity and the 1000 ton per day prototype mill. Costs would vary. The assumed new mill costs were used considerably among the mills for the expansions in the sample announced since 1971, and there was no need for additional pollution control identified for the announced new mills and capacity expansions since 1971.

Therefore, baseline estimates were not derived for this scenario. A fuel penalty of $.20 per million BTU was imposed for mills that were forced to employ low sulfur oil. This is considered to be the preferred control for most mills as they were currently burning oil and bark. If all mills are forced to increase the use of coal, it was assumed that stack gas scrubbers would be employed at costs indicated in Table 71.

For the prototype 1,000 ton per day new Kraft mill, a maximum annual cost penalty of $1.6 million or 1.5% of total annualized costs including capital and operating costs would be imposed if the mill were forced to burn low sulfur oil. In the unlikely event that mills were forced to shift to coal as a fuel, a mill may have an additional expense of $12.8 to $13.7 million or 7% increase in capital cost of the mill and an increase in total annualized cost of $4 million or a 3.7% increase if the mill were forced to use a stack gas scrubber along with locally available coal. As noted previously in most instances affected mills would not be required to reduce emissions to this extent in order to meet the proposed air quality and if these mills were, in fact, to be rebuilt under new NSD requirements siting, design and capacity alternatives would be available to significantly reduce if not eliminate these potential costs.

Nationwide, the costs of meeting the Senate and House proposed non-significant deterioration air quality increments appear to be minimal, less than a one percent increase in capital costs and approximately one percent increase in annual costs if 43% of total kraft pulping capacity including announced additions through 1978 were duplicated. These conclusions remain the same even if the mills were forced to burn coal. Table 72 summarizes the economic cost analysis assuming the capacity under study need only meet NSPS and was utilizing fuels actually burned.

7.2 Costs of compliance with BACT requirements

A separate provision under both the Senate and House non-significant deterioration proposals requires all new mills and capacity expansions to employ best available control technology (BACT) regardless of whether the mills will comply with the proposed air quality increments merely by meeting national new source performance standards (NSPS). The House BACT provision requires EPA to establish for national applicability more stringent new source performance standards than currently exist. For the purposes of analysis, the House BACT provision is conservatively interpreted as requiring either low sulfur oil (.3 lb. SO/million BTU), or stack gas scrubbers to be employed with high sulfur coal (.32 SO/million BTU) on all new mills as described in Table 71. The Senate provision is interpreted as a range of costs with the House requirements as an upper bound and EPA's current NSPS as a lower bound depending upon the case-by-case determinations made by a State.

Tables 73 and 74 are analyses of the cost impact of the BACT provisions on sample site capacity scenarios as compared to costs of just meeting the air quality increments. For additions through 1978, control costs take actual expansion size into account. As Table 71 indicates, there is a decrease in control costs per ton of capacity for larger additions due to economies of scale. For the other two sample site capacity scenarios, i.e., duplicating total sample site capacity and constructing a prototype 1,000 ton per day bleached Kraft mill at each sample site, control costs were based on the 1,000 ton per day prototype mill. In determining baseline costs, the capital and annual costs were based on the prototype 800 ton per day bleached Kraft mill described in section 7.1.

As can be seen in Tables 73 and 74, the capital cost of employing scrubbers with locally available coal, which is a worst case assuming all new additions burn coal, would only cause a 5% increase in industry capital requirements under the various scenarios. Annual costs (including operating and maintenance plus annualized capital costs) would increase by a maximum of only 2 to 4% depending upon whether new additions use lower sulfur oil or stack gas scrubbers with coal. When costs of control for the period 1971-78 were projected over the industry as a whole, as was done for the decade 1980-90 (see Table 75), annual costs would increase a maximum of 1.5% by 1990. Thus, the costs of even the most stringent control alternatives, nationwide, should not impose a significant burden on the industry.

7.3 Regional impacts

As discussed in Section 2, for the purposes of analysis the country was divided into six economic regions based upon U.S. Economic Census data. The alternative non-significant deterioration proposals affected the national economic regions quite differently. The regions with the principal impacts are presented in Tables 76 and 77. No new mills or capacity expansions since 1971 were required to alter fuel use or utilize additional control technology to comply with Senate or House increments. However, if total capacity at sample sites were duplicated under proposed Senate Class I and Class II increments, 16% of capacity in the mountain and Pacific region and 10% of capacity in the northeast would have been precluded if additional controls were not employed or fuel use altered. These numbers increase to 31% and 22% in each region, respectively, if it is assumed that coal meeting NSPS is burned at all mills in the sample. Under the proposed House Class I and Class II increments 24% of capacity in the mountain and Pacific region and 17% of capacity in the northeast would have been precluded if additional pollution control were not employed. These numbers increase to 35% and 24% in each region, respectively, if it is assumed that coal meeting NSPS is burned at all mills in the sample. Other regions of the country would have had no impact or minimal impacts on total sample mill capacity.

Almost all of this capacity could have been built in these regions if low sulfur oil or scrubbers were employed. The reasons for the variation in regional impact were the terrain characteristics in the northeast and mountain and Pacific regions. Capacity constraints that were identified under the Senate and House Class II increments occurred at mill sites in very hilly terrain with slopes greater than 6% within a 7 km radius around the mill. Among the sites in the sample, such terrain features were found exclusively in the northeast and mountain and Pacific regions. Siting will have to be more selective in these areas of the country. Nonetheless, low sulfur oil option for the impacted mills will not impose a significant cost (less than 5% of annual costs) and should allow most any economically sized mill to be built in terrain that is not extreme. Moreover, if all new capacity additions were required to employ control beyond NSPS, there would be much less distinction in impacts between regions. The House provisions requires EPA to determine what additional controls are required. The Senate provision leaves the question of additional controls up to the discretion of the individual State.

APPENDIX A: ALTERNATIVE POLICIES FOR THE PREVENTION OF SIGNIFICANT DETERIORATION AND RELATED POLICY ISSUES

As indicated in Table A1, the general approach to "significant deterioration" taken by EPA and by the Congressional Subcommittees is similar, although numeric limits and coverage vary considerably. In all cases, significant deterioration is prevented by establishing area classifications designed to correspond to the overall air quality intended for the area, and to reflect the amount of energy or industrial growth desired. These "classes"were established by the EPA regulations as follows:

Class I — Applies to areas in which practically any air quality deterioration would be considered significant, thus allowing little or no energy or industrial development.

Class II — Applies to areas in which deterioration that would normally accompany moderate, well-controlled growth would not be considered significant.

Class III — Applies to areas in which deterioration would be permitted to allow concentrated or very large scale energy or industrial development, as long as the national secondary ambient air quality standards are not exceeded.

It should be noted that a critical aspect of the EPA regulations is that all regions are initially designated as Class II, subject to redesignation as Class I or Class III by initiative at the State and local level. EPA anticipated that Class I designations would be made to protect existing clean air resources in areas such as national parks and wilderness areas, and the Class III redesignations would occur where State or local policies called for extensive industrial development, requiring increment air quality increases up to the national ambient air quality standards. To date, no known formal application has been made by communities, states or Federal land managers to reclassify regions to the Class I or Class III designations.

As indicated in Table A1, the Senate numerical limits for Class I and II areas are the same as EPA's. However, the Senate approach has several substantial differences:

The Senate version eliminates Class III designations.

The Senate requires the designation of some areas of national importance as mandatory Class I and other areas as discretionary Class I unless they are redesignated by agreement between the States and the Federal government. (See Table A2) .

The House also adopts the EPA approach but contains the following differences:

The House has fewer mandatory Class I designations than the Senate. (See Table A2).

The House version includes different numerical limits for the Class I and II increments, but the only significant difference is that the three-hour sulfur dioxide Class II increment is less than half of the EPA and Senate Class II increments. (See Table A1).

The House, like EPA, has a Class III but the allowed increment is half the NAAQS. In addition, a general air quality ceiling of 90% of the national primary ambient air quality standards or 100% of national secondary ambient air quality standards, whichever is more stringent, is imposed.

The other important differences among the alternative proposals are summarized in Table A3. One of these is related to the control technology a new power plant must install regardless of the applicable air quality increment. While EPA's significant deterioration regulations require compliance with New Source Performance Standards (NSPS) or State regulations where they are more stringent than NSPS, the Senate requires the maximum degree of control technology achievable as determined on a case-by-case basis. Similarly, the House proposal requires the best available control technology which (taking into account cost, other environmental impacts, and energy requirements) the Administrator of EPA has determined to be adequately demonstrated.

Another important difference is that the House proposal provides for limitations on the allowable stack height for new power plants. Specifically, the House proposal restricts the allowable stack height to two and a half times the height of the power plant unless the source can demonstrate that a greater height is necessary to insure that emissions from the source do not cause excessive concentrations of any air pollutant in the immediate vicinity of the source as a result of atmospheric downwash, eddies, and wakes.

APPENDIX B: AIR QUALITY IMPACT OF NEW KRAFT PULP MILLS

To assist in determining the impact of the House and Senate significant deterioration amendments on Kraft pulp mills, the air quality impact of two mill sizes was examined under various stack height and terrain conditions. The results are intended to be used as a screening procedure only; specific modeling must be performed for individual mills to determine if a restriction, in fact, exists.

Modeling assumptions

EPA's Single Source Model (CRSTER) was used with meteorological data from sites in the northeast, northwest, and the southeast for the cases where terrain is below the physical stack height. The meteorological data was selected to be representative of sites where pulp mills would be located. At longer downwind distances for such cases, the maximum concentrations predicted by the Single Source Model were for very stable and/or calm wind situations. Since plume travel time is not accounted for in the model, these concentrations are not realistic and should not be used for the Class I intrusion analysis. The conditions used in deriving Figure 3 are considered much more realistic for longer downwind distance calculations.

EPA's Valley Model, which uses assumed worst case meteorological conditions (E stability; u=2.5 mps), was used for cases where the plume impinges on elevator terrain (135 meters higher than the stack for a 1000 ton per day mill).

For the case where terrain features are above the physical stack height but less than the height where plume impingement was assumed (135 meters higher than the stack for the 1000 tpd mill) a different modeling approach was used.

A PTMAX run was made with terrain exceeding the stack by 150 feet to see which meteorological conditions resulted in the highest concentrations. The results, which are based on the stack parameters for the 1000 tpd mill, were as follows: plotted as a function of distance (Figure 1). For comparison purposes, the concentrations for plume impingement are also shown (upper lines in Figure 1).

The maximum concentrations for C and D stability in the table above and in Figure 1 are overestimates, since under unstable and neutral conditions, the plume is normally assumed to follow the terrain (see Figure 2). Thus, there was little point in examining cases where stack height exceeded terrain by more than 150 feet for the C and D stability cases.

The stack parameters used in the diffusion modeling are summarized in Table 2. A normalized emission rate of 100 lb/hr was used in most of the model runs, so that the predicted concentrations must be adjusted, as described below, to be consistent with the specific new mills being considered. To simplify the modeling, all emissions were assumed to be released through single stack. Annual averages were not calculated, since the short term increments are expected to be controlling.

Separate model runs were not made for the case where a scrubber is used to control the power boiler emissions, since there is little difference in the stack parameters for the scrubber and non-scrubber cases. Due to the reheating effect of the recovery furnace flue gas, the final stack gas temperature when a scrubber is used is approximately 280ºF, as compared to 350ºF for the non-scrubber case. An average of these two temperatures was used in the modeling.

Emissions

Table 3 lists the emissions for a 1000 tpd mill for various fuel and power generation conditions. BACT for sulfur dioxide from the recovery furnace is assumed to be 30 ppm (1.6 lbs per ton or 67 lbs per hour). Emissions for the coal and oil combustion cases were calculated directly. from existing or tentative NSPS for Kraft mills and steam generators, except that for particulate matter from oil combustion, the AP42 uncontrolled emission factor was used. Where bark is being burned, the numbers on page B2 of ERT's report on Kraft mills were used as a basis, but were adjusted to account for lower sulfur dioxide emission from the recovery furnace and lower particulate emissions from oil combustion, in accordance with the assumptions discussed above.

Results and use of data

The highest concentrations predicted with the Single Source CRSTER Model (Tables 4and 5) adjusted for the highest emission rate in Table 2 (1127 lb SO per hour), were 48 ug/m a, 24 hour average; and 130 ug/m s, 3 hour average. Thus, when terrain does not exceed the physical stack height, the House and Senate Class II increments would both permit mills in excess of 2000 tpd. In this terrain situation, the 1000 ton per day for the highest emission rate would not exceed the Class I increment in an adjacent area if it were located more than 16 km under the Senate proposal or 9 km under the House proposal.

Where the plume impinges on elevated terrain (Figure 4), a 1000 tpd mill would violate the Senate Class II increments out to 5 km (24 hour increment is controlling) and the House Class increment out to 6 km (3 hour increment is controlling) while the Senate Class I increments would be exceeded out to 45 km (24 hour increment is controlling) and the House Class I would be exceeded out to 32 km (3 hour increment is controlling). The plume impingement curves of Figure 4 are valid when the terrain exceeds the stack height by 135 meters for the 1000 tpd mill or 100 meters for the 400 tpd mill (based on plume rise under E stability, u=2.5 mps). Since E stability with constant direction winds was assumed to persist for no more than 12 hours, the curves of Figure 1 are not valid beyond 81 km for 3 hour average (9 hours of travel time plus 3 hours of impingement) or 54 km for 24 hour average (6 hours of travel time plus 6 hours of impingement).

The following screening procedure was used to determine whether an individual mill would be restricted by either the Class II or III (House) increments or an adjacent Class I area.

1. Use the runs that correspond most closely to the individual mill (i.e., size and area of the country).

2. Using topographical maps, find the minimum emission height (stack height minus maximum terrain height) at various downward distances.

3. For each downward distance, find the normalized concentration from the appropriate figure or table, as follows:

A. Terrain up to the height of the stack.

Tables 4 and 5 contain results that should be used for non-impingement cases for detaining maximum concentrations (Class II analysis only). For Class I intrusion analyses and the interaction of two sources at distances greater than the location of the maximum concentration, Figure 3 should be used.

B. Terrain between 0+135 meters higher than top of stack.

Figure 1 curve c should be used.

C. Plume impingement (terrain greater than 135 meters above top stack) .

Figure 4 should be used out to the limits discussed above. For greater distances, Figure 3 should be used.

4. Adjust normalized concentration according to the ratio of the actual emissions, [i.e., X1=Xn (Q1/Qn), where the subscript 1 denotes the individual mill, n denotes the normalized mill, X is concentration, and Q is emission rate in consistent units for both i and n.

5. Compare resulting concentrations with Senate and House increments applicable at each downwind distance. Where interaction of two sources is being examined, consider only terrain that is colinear with the two sources, adding concentrations at appropriate downwind distances.