
                      UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                          SECTOR POLICIES AND PROGRAMS DIVISION
                       OFFICE OF AIR QUALITY PLANNING AND STANDARDS
                               OFFICE OF AIR AND RADIATION


October 23, 2011

MEMORANDUM

SUBJECT:	Summary of Kraft Condensate Control Technology Review
	
FROM:	John Bradfield
		Kelley Spence	
	
TO:		Docket EPA-HQ-OAR-2007-0544


   This memorandum summarizes the technology review of kraft pulping condensate control processes in pulp mills subject to 40 CFR part 63, subpart S analyzed for the risk and technology review (RTR) of the national emission standards for hazardous air pollutants (NESHAP) for the pulp and paper industry. 

   I. Background
   
   Kraft pulping condensates are hazardous air pollutant (HAP)-containing liquids that are condensed from pulping system vent streams. These HAP-containing liquids result from the contact of organic compounds in the pulping process with water or steam that condenses and is collected. Prior to the subpart S standards (63 FR 18504, April 15, 1998), condensates could be discharged into open collection systems and transported to either the onsite industrial wastewater treatment plant or a publicly owned wastewater treatment plant. Subpart S required mills to either (1) recycle the condensates back to equipment that meet the control standards for pulping system vents (low-volume high concentration [LVHC], high-volume low concentration [HVLC]), (2) treat the condensates to reduce or destroy the HAP by at least 92 percent by weight, (3) treat the condensates to remove a specified amount of HAPs (92% of 11.1 pounds per oven-dried ton of pulp [lb/ODTP] or at least 10.2 lb/ODTPat mills performing bleaching and 92% of 7.2 lb/ODTP or 6.6 lb/ODTP at mills without bleaching), or (4) treat the condensates to meet a specified HAP concentration at the control device outlet (330 parts per million by weight [ppmw] at mills performing bleaching or 210 ppmw at mills without bleaching). All condensate streams must be collected, including those from:
      
   1. The digester system  -  blow heat accumulator overflow and secondary condensers
   2. The turpentine recovery system (if present)  -  decanter underflows
   3. The evaporator  -  foul condensates from the weak liquor feed and vacuum systems only
   4. Non-condensable gas (NCG) systems  -  condensates from the LVHC and HVLC collection systems and any steam strippers
        
   Because of other process considerations, mills generally chose the 92 percent control options for compliance demonstration for kraft condensates rather than recycling. Only five mills practice recycling to control kraft condensates, and two mills utilize both recycling and steam stripping to control emissions. Only 4 mills indicated that they use the ppmw alternative. Consequently, the focus of this technology review is on the control efficiencies of wastewater treatment systems and steam stripping/incineration.

   I. HAP Emissions From Kraft Condensates

    In their responses to the 2011 EPA information collection request (ICR), a majority of mills indicated the quantity of HAP in kraft condensates collected from the sources identified in 40 CFR 63.446. The remainder only indicated the percentage of HAP in kraft condensates collected and treated.  The average kraft mill collected 13.4 lb HAP condensates per ODTP produced.  The production volume of pulp produced in 2009 reported was approximately 40 million ODT.  The average percentage removal of kraft condensates was 96.2 percent.  Consequently, the approximate HAP emissions from kraft condensates collected were 10,000 tons.  The portion of these emissions that is hardpiped to wastewater treatment facilities is available for improved kraft condensate control through biological treatment. The portion of these kraft condensates that is steam stripped then incinerated is available for improved control through improved steam stripping efficiency. Mills had the option of simply reporting the percentage of kraft condensates treated in their ICR response (one of the subpart S compliance options), so collection volumes were not received from every mill.

   II. Developments in Practices, Processes and Control Technologies

   A. Steam Stripping/Incineration

The incineration requirements for stripper off gases (SOG), the HAP eluting from a steam stripper, are set forth in 40 CFR 63.434(d) and require the use of a thermal oxidizer that operates with at least a 98 percent VOC destruction efficiency, a power boiler, or a pulp mill production combustion device (e.g. lime kiln, recovery boiler). Boilers or production combustion devices operating under the specifications of the standard are presumed to destroy all the organic HAP present. EPA review of the compliance test information provided via the ICR conducted in conjunction with the subpart S RTR indicated that mills using oxidizers had destruction efficiencies in excess of 99 percent. We did not find a control technology that improved on this level of control that can be applied to this source. (EPA 2003) Consequently, we have determined that incineration of SOG remains the best technology to control this source. Additionally, since the recycling approach to kraft condensate control feeds emissions into either an LVHC or HVLC system where NCG are incinerated, no further technology review was determined to be necessary for mills that chose this regulatory option. 

The ICR results showed 52 mills that practiced steam stripper control of kraft condensates, 9 of which also used biological treatment. When the original subpart S was promulgated, it was determined that the steam stripper efficiency for the removal of methanol from kraft condensates was 92 percent (See Docket A-92-40, IV-B-10). The range of stripper efficiencies for methanol at the 32 mills studied was -9 percent to 99.5 percent. The maximum achievable control technology (MACT) floor calculations from that dataset returned a floor level of 92 percent, which was codified into the subpart S standard. Since methanol was determined to be the most difficult of the HAP present to be steam stripped from the kraft condensate stream, it was presumed that a measurement of efficiency for the removal of methanol was a conservative estimate of HAP removal. 

EPA performed a comparison analysis of (1) the steam stripper collection efficiency data collected during the development of the original subpart S rule and (2) the data collected for the 2011 ICR. Since 100 percent of the industry responded to the ICR, for purposes of comparison, the analysis included a broader dataset of the stripper data than that used to develop the original standard (See Docket A-92-40, IV-B-9 and IV-J-32) and employed an upper predictive limit (UPL) statistical analysis that reflects current approaches used for statistical regulatory analysis. The UPL statistical analysis of the current performance of the best performing mills demonstrating compliance using the methanol removal efficiency option actually showed a slightly lower (98.6 percent vs. 96.1 percent) removal efficiency for the mills in the ICR database when compared to the older database. Further, the removal efficiencies in compliance demonstration tests ranged from 92.1 percent to 98.6 percent control, a higher average but a lower peak control than the earlier dataset, so this statistical outcome is understandable. Additionally, since a much broader database of mills was available for the analysis, the current dataset is much more reflective of the actual application of this technology. Consequently we have determined that, on average, there have been no general technological improvements in the demonstrated performance capabilities of steam strippers used for kraft condensate compliance to the subpart S standard.

However, we did not limit our technology analysis strictly to the expected or average control capabilities of the steam stripper process, as the Clean Air Act (CAA) requires the agency to also review practices and processes.  We also performed a work practice analysis.  To review work practices, we chose to evaluate the entire dataset range to determine whether all or virtually all of the mills employing steam strippers were demonstrating control at levels exceeding the standard, in practice. Though we requested additional voluntary comment on practices in the ICR, we received a limited response since there was a large volume of mandatory data requested in the ICR and a short response period. Though we received no specific information about quality, process control, and work practice improvements, we found we could evaluate the data received in a general way using the methanol 92 percent control capture reference. The requirements were written as performance-based numerical limits (i.e. 92 percent capture).  Since mills were left to find their own best practice to meet the performance standard, a strict analysis of numerical compliance levels of actual results in meeting the standard was found to be appropriate. The dataset of performance values found every mill exceeded the 92 percent capture level. One mill only reported a steam stripping capture efficiency of 92.1 percent, but for this mill, steam stripping was only a backup treatment option.

There are 97 mills currently operating in the kraft pulping portion of the category. Of these, 4 exceeded the 92 percent minimum control but did not exceed 93 percent. Further, 15 did not exceed 94 percent control, 28 did not exceed 95 percent control, 41 did not exceed 96 percent control, 54 did not exceed 97 percent control, and 66 did not exceed 98 percent control. An additional compliance demonstration allows mills to show that, using either steam stripping or biological treatment; a minimum quantity of condensate was collected and removed. For bleached mills, 10.2 lbs of HAP/ODT of pulp produced is the minimum equivalent to 92 percent control, and for unbleached mills it is 6.6 lb/ODTP.  We determined that the proportional increase percentage for each step between 92 percent and 100 percent, (12.5 percent for 92 percent to 93 percent, 25 percent from 92 percent to 94 percent, etc.) could be applied to the baseline minimum quantities of, respectively, 10.2 and 6.6 lb/ODTP for comparison purposes. Thus, the dataset was developed for purposes of evaluating regulatory options. See Table 1.
   
   Example calculation:
   
   1%100%-92%=12.5%

Table 1. Number of Kraft Mills Meeting Kraft Condensate Removal Levels, by Percentage
                                    Control
                                      92%
                                      93%
                                      94%
                                      95%
                                      96%
                                      97%
                                      98%
                             Standard improvement
                                      0%
                                     12.5%
                                      25%
                                     37.5%
                                      50%
                                     62.5%
                                      75%
                      No. mills meeting improved standard
                                      97
                                      93
                                      82
                                      69
                                      56
                                      43
                                      31

   B. Wastewater Treatment Control (Hardpiping Option)
   
   Wastewater treatment (WWT) served as an intersection point in the original cluster rule (i.e., subpart S and effluent guidelines), the point where controls for air emissions integrate, inter-grade, and overlap and where, potentially, engineering judgments are made so mills meet both the subpart S HAP emission limits and effluent control standards. As noted, mills can use biological treatment to remove HAP generated as kraft condensates, the so-called `hardpiping' option. Hardpiping requires condensates to be discharged from the pulping process below the liquid surface of a biological treatment system where they are treated through biological action. 
   
   There are three general technology approaches to wastewater treatment in the pulp and paper industry: (1) aerated stabilization basins (ASB), (2) activated sludge treatment systems (AST) and (3) pure oxygen activated sludge systems (e.g., UNOX). Additionally, several mills use an anaerobic treatment step in their process of wastewater treatment. The hardpiping option requires compliance demonstrations, which show that the WWT system (via biological treatment) is at least as effective as methanol steam stripping at removing HAP, i.e., 92 percent removal efficiency. The measure of removal requires a `total HAP' reduction approach, where a surrogate for total HAP is a measurement of reduction volumes for four HAP: methanol, acetaldehyde, methyl ethyl ketone (present on the HAP list in 1998) and propionaldehyde. All three systems have been successful at reducing kraft condensates by efficiencies greater than 92 percent, and none of the systems have demonstrated greater efficiency at this type of control, though the AST and UNOX systems have potential advantages over ASB for purposes of meeting effluent guidelines and total reduced sulfur (TRS) emission reductions from WWT. Consequently, even though these two systems are typically more expensive to operate than ASB, they are still in common use at pulp and paper mills.
   
   All the compliance demonstrations for wastewater treatment are in the same format as stripping/incineration. Mills can show 92 percent reductions or show an emissions reduction against the comparable 10.2 lb/ODTP bleached and the 6.6 lb/ODTP unbleached standards. When a comparison of steam stripping control versus biological control was made, there were no significant differences between the two principal options for eliminating kraft condensate emissions. The average percentage reduction numbers for steam stripping were slightly lower than WWT (95 percent to 96 percent), but the lb/ODTP emissions reduction numbers were better for the steam strippers. All the values differed by less than a standard deviation of their internal variability.
   
   A process analysis was also performed on the WWT sources similar to the analysis above. The results for the kraft condensate removal level comparison for both steam stripping and biological control are noted in Table 1.

   C. RBLC Review
   
   Review of EPA's RACT/BACT/LAER Clearinghouse (RBLC) indicated that no mills have received best available control technology (BACT) determinations for kraft condensate control under the new source review (NSR) permitting program. Typically, these determinations involve some type of process change (e.g., use of revamped or new systems with recent control innovations). Though there were many changes at mills under subpart S implementation requirements for kraft condensate control, none of them required NSR permits. One mill that was identified in the RBLC did undergo a BACT-PSD determination for their wastewater system in 2001. However, that mill recycles its kraft condensates, and the BACT review covered other sources and did not require the use of technology different than the three typical systems in use for kraft condensate biological control.

   D. New Controls for Consideration
   
   Even though we found no new applied technologies in our RBLC review, we did identify a number of wastewater control technologies that are not currently used in the pulp and paper industry. There are several other very effective control systems that have the ability to remove potentially volatile HAP from water streams like those containing kraft condensates. Other technologies reviewed included carbon absorption technology, chemical oxidation and membrane separation. A screening review indicated that none of these technologies had HAP removal performances better than the very best of the three current general wastewater treatment approaches in the industry. Since the costs associated with each of these potential technologies indicated control costs would range above the highest cost to be considered in the technology review of in-place technologies, they were not considered as optional technologies and were not considered further.

   I. Residual Risk from Kraft Condensates

   The residual risks from sources with emissions from kraft condensates come primarily from wastewater. The wastewater emissions that generate some level of risk that may have come from kraft condensate sources are primarily acetaldehyde. There were no mills that exceeded a chronic non-cancer HI of 1. The highest acute non-cancer, with a hazard quotient (reference exposure level) (HQ [REL]) of 20 was acetaldehyde from wastewater. Also, 6 other mills had acute non-cancer HQ (REL) above 1 for one of either acetaldehyde, formaldehyde or methanol, emissions that can be generated from kraft condensates. The highest HQ mill did not use biological control for emissions from kraft condensates. Improvements in control of kraft condensates would not impact HQ (REL) risk from this source. It is not certain about the degree to which emissions from kraft condensates contributed to this risk since other sources of these compounds can be found in wastewater streams in paper mills. Three of the other six mills with and HQ (REL) above 1 used biological treatment for their kraft condensates. Options that reduce emissions from kraft condensates from mills that use biological control could reduce risk exposure from those sources.  

   II. Regulatory Options

   Our technology review of the RBLC, as noted above, did not yield any information about new applied technologies that could become the basis for regulatory options. The next consideration was an analysis of the ICR database to determine if, within the current control technologies, there were mills that had demonstrated the ability to remove greater HAP quantities than the requisite 92 percent control (or equivalent demonstrations). Then we analyzed the data recovered to see if there was a predominant portion of the industry that had performed significantly better than the 92 percent minimum control level. We found in both cases that there were mills performing better than the 92 percent minimum and that the great majorities were performing significantly better. Consequently, in the first set of regulatory options developed, we developed an incremental scale of improvement over the minimum 92 percent control. We decided on an evaluation scale set up by percentage increments from 93 percent to 98 percent. We did not take the comparison beyond 98 percent because that level was determined to be at the limit of control efficiency for one the major control techniques, steam stripping, and it was equivalent to the control level required for non-condensable gases ducted to controls from LVHC and HVLC sources in 40 CFR 63.443(d).

   Before an incremental comparison could be made, first an equivalency needed to be established between the different compliance and demonstration options.  There are four general approaches: the two primary technologies used to control kraft condensates--biological treatment and steam stripping/incineration, demonstrating through a  generic option based on control device exit parameters (in ppmw), and through recycling. Prior to steam stripping, mills can segregate the stream and control with different compliance demonstrations based on that segregation.  Then, through steam stripping or biological control, each segregated stream or all streams must demonstrate control through (1) 92 percent or greater collection or control; (2) a 10.2 lb HAP/ODTP or greater reduction for bleached pulp sources or (3) a 6.6 lb HAP/ODTP or greater reduction for unbleached pulp sources.

   Table 2 provides the equivalency measures used for all the compliance and demonstration options. Since the recycle option requires the return of condensates to a source controlled by the incineration standards listed in 40 CFR 63.443(d), recycled condensates are presumed to always be controlled at a level of 98+ percent. The compliance demonstration equivalents to, respectively, 93, 94, 95, 96, 97 and 98 percent control became the 6 regulatory options developed on the basis of improvements in existing technology codified into the original rule. See Table 3.

Table 2. Compliance and Demonstration Option Equivalents, by Increment of Improvement
                        Control by biological treatment
                     or steam stripping then incineration
                                General control
                                    device
                            General control device
                                Control through
                                    recycle
                                By % reduction
                            (% control improvement)
                        Bleached mass/volume reduction 
                                  lb HAP/ODTP
                       Unbleached mass/volume reduction 
                                 lb HAP/ ODTP
                                   Bleached 
                                     pulp
                                     ppmw
                                  Unbleached 
                                     pulp
                                      ppmw
                                       
                                     92%*
                                     10.2*
                                     6.6*
                                     330*
                                     210*
                                      N/A
                                  93% (12.5%)
                                     11.5
                                      7.4
                                      289
                                      184
                                      N/A
                                   94% (25%)
                                     12.8
                                      8.3
                                      248
                                      158
                                      N/A
                                  95% (37.5%)
                                     14.0
                                      9.1
                                      206
                                      131
                                      N/A
                                   96% (50%)
                                     15.3
                                      9.9
                                      165
                                      105
                                      N/A
                                  97% (62.5%)
                                     16.6
                                     10.7
                                      124
                                      79
                                      N/A
                                   98% (75%)
                                     17.9
                                     11.6
                                      83
                                      53
                                  Presumed**
*Current subpart S standard.
**Following control per 40 CFR 63.443(d).

                          Table 3. Regulatory Options
                                    Option
                                  Requirement
                                       1
Improve control efficiency from a minimum of 92% (or equivalent) to a minimum of 93%
                                       2
Improve control efficiency from a minimum of 92% (or equivalent)  to a minimum of 94%
                                       3
Improve control efficiency from a minimum of 92% (or equivalent)  to a minimum of 95%
                                       4
Improve control efficiency from a minimum of 92% (or equivalent)  to a minimum of 96%
                                       5
Improve control efficiency from a minimum of 92% (or equivalent) to a minimum of 97%
                                       6
Improve control efficiency from a minimum of 92% (or equivalent)  to a minimum of 98%

   III. Costs

   Table 4 shows the estimated costs associated with each regulatory option referenced above.  The following general presumptions were made regarding costs developed for each option.  See T. Holloway, K. Hanks and C. Gooden, RTI, Memo to Docket EPA-HQ-OAR-2007-0544, Cost and Environmental and Energy Impacts for Subpart S Risk and Technology Review (Reference 5) for detailed information.

For mills demonstrating compliance with biological control:
   * Cost of upgrade required is proportional to improvement needed; i.e., 92 > 93 = Current Size * 93/92, etc.
   * Operation and maintenance cost ($/yr) = 0.2546 x MGD[0.6969]  x 1,000,000.
   * Capital costs follow the following equation: Capital cost ($ million) = 5.9749 x MGD[0.8228]  x 1,000,000
   * Capital recovery factor: 0.094  -  based on 7 percent interest over 20 years.
   * 57 to 72 percent of the increased operation and maintenance costs will be for energy, primarily for pumping and aeration.

For mills demonstrating compliance with steam stripper/stripper off gas control:
   * Presumptive capital cost, new unit: $5,335,000
   * Presumption for efficiency improvement 2 percent or less: rebuild unit, capital cost (1/2) of new.
   * Presumption for efficiency improvement 3 to 6 percent: install new unit.
   * Presumptive annual cost increase: $0.  Mills using steam stripper compliance will already be paying the costs of operation and maintenance, and the operation and maintenance on a new or rebuilt unit will be equal or less.  Similarly, energy costs will not change or be less since a new or rebuilt unit will be equivalent or better in its energy efficiency.

                      Table 4. Regulatory Costs by Option
                                    Option
                                    Minimum
                                  collection
                                  efficiency
                                      or
                                  equivalent
                      No. of impacted mills of total (97)
           Nationwide potential incremental emission reduction, tpy
                                Total capital,
                                   $million
                         Total annualized, $million/yr
                                  Cost eff.,
                                     $/ton
                         Incre-mental cost eff., $/ton
                                      --
                                  92% - base
                                                                              0
                                                                              0
                                                                              0
                                                                              0
                                                                              0
                                                                              0
                                       1
                                92% -->93% 93%
                                                                              4
                                               2,046                     2,046 
                                                                           $8.7
                                                                          $0.99
                                                                           $485
                                                                             --
                                       2
                                  92% -->94%
                                                                             15
                                          4,092                          4,092 
                                                                            $36
                                                                           $4.1
                                                                         $1,010
                                     $1,512                            $5,,380 
                                       3
                               92% -->95%  -->93
                                                                             28
                                          6,138                          6,138 
                                                                            $79
                                                                           $9.0
                                                                         $1,460
                                       $2,390                         1$5,,240 
                                       4
                              92% -->96%   -->94
                                                                             41
                                          8,183                          8,183 
                                                                           $142
                                                                            $16
                                                                         $1,980
                                       $3,421                         3$2,,744 
                                       5
                                       
                                  92% -->97%
                                                                             54
                                          10,229                        10,229 
                                                                           $218
                                                                            $25
                                                                         $2,420
                                        $4,398                         52,,177 
                                       6
                                  92% -->98%
                                                                             66
                                          12,275                        12,275 
                                                                           $297
                                                                            $34
                                                                         $2,700
                                         $5,366                         67,668 

   IV. References

   1. G. Smook.  Handbook for Pulp & Paper Technologists.  3[rd] Edition. Angus Wilde Publications Inc., Bellingham WA.  2002. 

   2. Pulp and Paper ICR.  2011

   3. U.S. EPA.  OAQPS Control Cost Manual. 2003.

   4. U.S. EPA.  Technical Report: Construction Costs for Municipal Wastewater Treatment Plants: 1973-1978. EPA/430/9-80-003. Washington DC.  1980.

   5. T. Holloway, K. Hanks, and C. Gooden, RTI.  Memo to J. Bradfield and B. Schrock.  Cost and Environmental and Energy Impacts for Subpart S Risk and Technology Review.  2011.
