Biogas Cogeneration System Sizing and Payback Based on Weekly Patterns of Anaerobic Digestion and Biosolids Dryer Operation

Transcription

1 Biogas Cogeneration System Sizing and Payback Based on Weekly Patterns of Anaerobic Digestion and Biosolids Dryer Operation ABSTRACT J.C. Kabouris, PhD, PE 1 ; CH2M HILL R. Forbes, PE, CH2M HILL T.G. Shea, PhD, PE, BCEE, CH2M HILL M. Engelmann, PE, Pinellas County Utilities J. Dulaney, Pinellas County Utilities 1 CH2M HILL 4350 W. Cypress St., #600 Tampa, FL John.Kabouris@ch2m.com Pinellas County Utilities (PCU) is implementing digestion improvements at its 33 mgd South Cross Bayou Water Reclamation Facility (SCBWRF) that will allow the codigestion of sludge and fats, oil and grease (FOG) feedstocks. Based on the expected high biogas production due to the codigestion of FOG and also due to a limit in the biogas utilization by the SCBWRF dryer/pelletizer, significant biogas quantities are expected to remain available after the pelletizer s biogas needs are met. The economic feasibility of implementing combined heat and power (CHP) generation at SCBWRF was analyzed in order to maximize lifecycle benefits and avoid energy wastage through biogas flaring. Analysis of various gas production scenarios considered the amount of available FOG as well as the weekly pattern of biogas production and consumption, presuming that the FOG would be fed to the digesters only during weekdays when the pelletizer would also be in operation. Internal combustion engine (ICE) based cogeneration provided the maximum lifecycle benefit among cogeneration alternatives. Based on the target biogas scenario, the payback period of a 1,400 kw (electrical) ICE cogeneration system and associated biogas treatment was estimated to be 7.3 years, with a projected 20-year lifecycle benefit of $5,400,000. The sensitivity of these benefits to key assumptions was assessed by varying uncertain technical and financial parameters. KEYWORDS Cogeneration, dryer, dewatered FOG, acid-methane thermophilic digestion, sensitivity analysis. INTRODUCTION Pinellas County Utilities (PCU), Florida operates the South Cross Bayou Water Reclamation Facility (SCBWRF), rated at 33 mgd. The biogas produced at SCBWRF is used by the on-site biosolids drying/pelletizing process and reduces the quantity of natural gas (NG) that otherwise would be required. The PCU is presently trucking belt filter press dewatered biosolids at about percent solids concentration from its 9 mgd William E. Dunn WRF (WEDWRF) to the SCBWRF and processing it directly in the drying/pelletizing facilities.

2 Kabouris et al (2008) evaluated the option of converting the anaerobic digestion operation to a three-phase mode of digestion operation (acid mesophilic / methane thermophilic / non-heated storage), including the codigestion of five feedstocks: SCBWRF thickened solids, WEDWRF cake, fats, oil and grease (FOG) dewatered by the PCU using polymer, and lime-dewatered FOG from a private hauler. PCU proceeded with the design of digestion upgrades and the conceptual design of heat and power cogeneration, with the objective of utilizing the energy from biogas available when the dryer is not operating. The cogeneration analysis is complicated by the many interactions among the various considered system components (digestion, dryer, cogeneration, boiler, flares) as depicted in Figure 1, and by the time-varying availability of digester gas. Most of the gas production is expected to occur during weekdays, when the FOG will be accepted and fed to the digesters. The dryer will also operate mostly during weekdays, so less biogas will likely be available for cogeneration during weekdays. Therefore, to produce realistic projections, a mass and energy balance approach for the digesters, dryer, and cogeneration system was constructed for both weekday and weekend conditions. The mass balance incorporated the results of laboratory scale batch and semicontinuous digestion for the various feedstock combinations, as described in Kabouris et al. (2008). Figure 1 Evaluated SCDWRF System Natural Gas (Back-up) Boiler Flares Utilized Waste Heat Wasted Biogas Fuel Digestion Biogas Fuel Biogas Fuel Dryer Unutilized Dryer Waste Heat Utilized Waste Heat Biogas Fuel Utilized Dryer Waste Heat Natural Gas Unutilized Waste Heat Cogeneration Natural Gas Electrical Power Generation This paper describes the process by which the biogas cogeneration system selection was justified based on projections of financial payback under a range of operating cases and variable digester gas production and availability. It includes sensitivity analysis results on the payback periods

3 based on the price of purchased electricity and natural gas, and quantifies the effects of operating the cogeneration on biogas with and without natural gas supplementation. ESTIMATION OF BIOGAS AVAILABLE FOR COGENERATION Conceptual Modeling of Weekly Biogas Generation The supplemental digestion feedstocks (polymer FOG, lime FOG, and WEDWRF cake) are expected to be fed to the digestion process only during weekdays. As a result, the gas production from the egg-shaped digesters (ESDs) is expected to be higher during weekdays than during weekends. In addition, the SCBWRF pelletizer typically also operates only during weekdays. It is therefore important to estimate the distribution of biogas production during weekdays versus weekends as well as the associated biogas consumption in the pelletizer. In this manner the cogeneration engines can be sized with sufficient peak and turndown capacity to optimize the storage and utilization of biogas during both weekdays and weekends, to minimize the likelihood of biogas wastage through flaring. A conceptual model for the weekly pattern of biogas generation was developed to facilitate the analysis and decision making. During the laboratory-scale acid-methane digestion testing of SCBWRF sludge and polymer FOG (Kabouris et al., 2008), it was observed that the gas production from the methane phase increased almost linearly following the introduction of the feed stream consisting of the acid reactor effluent. The transient period from the onset of the feeding to achievement of a steady-state plateau in methane phase biogas production lasted approximately 6 days. This corresponds to a biogas variation of approximately 16.7 percent of peak production per day. Therefore, total weekend biogas production rate is expected to be up to 33 percent lower than weekday production. The average biogas reduction over the weekend is assumed to be approximately half of the maximum, or up to about 16.7 percent. To simplify the calculations, a stepwise biogas production profile was implemented, with a reduction of 16.7 percent between weekday and weekend biogas production. The pelletizer typically operates an average of approximately 5 days per week, but this period may be longer or shorter depending on the weekly biosolids load to the dryer. To further simplify the analysis, it was also assumed that the high biogas production period would coincide with the dryer s operation period, because both dryer operation and FOG deliveries typically occur 5 days per week. This conceptual modeling approach was considered sufficient for the purposes of cogeneration feasibility analysis and preliminary system sizing. The actual biogas production will vary on a daily basis because of variability in the amounts of SCBWRF sludge, WEDWRF cake, and FOG quantities fed, as well as variability in operating conditions. Biogas Available for Cogeneration A mass and energy balance approach, similar to the approach used to derive the expected biogas quantities in the digestion upgrades PCU project (Kabouris et al., 2008), was used to estimate the biogas quantities that could become available for cogeneration following the digestion upgrades.

4 All calculations were performed based on the selected future operational scenario of three-phase digestion. The projected gas production is strongly related to the amount of FOG that will be available for digestion because of the very high biogas generation potential associated with the digestion of FOG. Due to digestion capacity and future availability uncertainty, the number of 15,000 wetpound truckloads of lime FOG delivered each weekday and processed in the SCBWRF digestion facilities following their upgrades is also uncertain. Accordingly, the analysis considered the sensitivity of the projected benefits based on one, two, or three such truckloads of lime FOG per weekday. In addition to the uncertainties in the biogas production and methane content, uncertainty also exists regarding the rate of biogas consumption by the dryer. According to information obtained from Andritz, the SCBWRF rotary dryer s manufacturer, up to 90 percent of the combined natural gas and biogas fuel blend for the dryer can be biogas. This assumption has been used in the analysis and design of the digestion upgrades (Kabouris et al., 2009) and was considered as one possible scenario. This assumption of high biogas utilization in the dryer results in low surplus biogas available for cogeneration, but is an unproven operational conditions, and PCU staff felt that there are constrains in the dryer s design parameters that could prevent it from having such a high biogas content in the dryer s fuel. To cover the full range of possibly available biogas for cogeneration, a second scenario was therefore considered, based on the historically low biogas utilization by the dryer. Under that scenario, it was assumed that only 35 percent of the combined natural gas and biogas fuel blend could be biogas, which is also the biogas percentage originally used by Andritz in the SCBWRF dryer design calculations. As a result of the low biogas consumption in the dryer, this scenario is associated with the availability of a large amount of surplus biogas and methane gas for beneficial utilization in cogeneration. Finally, the possibility of target and low digestibility for the thickened waste activated sludge (TWAS) was also considered, based on ultimate volatile solids reduction (VSR) values of 45 and 23 percent, respectively. The low digestibility TWAS was considered in conjunction with one lime FOG truckload per day, to estimate a minimum rate of biogas production. More specifically, the following eight cases were selected from the large number of possible combinations and analyzed using mass and energy balance calculations. A. Based on design biogas utilization by the dryer (35 percent of combined natural gas and biogas fuel blend being biogas; maximum biogas amount available for cogeneration). 1. One lime FOG truckload per weekday with WEDWRF cake digestion. 2. One lime FOG truckloads per weekday without cake digestion. 3. Two lime FOG truckloads per weekday with cake digestion. 4. Three lime FOG truckloads per weekday with cake digestion. 5. One lime FOG truckload per weekday with WEDWRF cake digestion and low TWAS digestibility. B. Based on high biogas utilization by the dryer (up to 90 percent of combined natural gas and biogas fuel blend being biogas; reduced biogas amount available for cogeneration). 1. One lime FOG truckload per weekday with WEDWRF cake digestion. 2. Three lime FOG truckloads per weekday with WEDWRF cake digestion.

5 3. One lime FOG truckload per weekday with WEDWRF cake digestion and low TWAS digestibility. Case A4 is the target scenario, representative of the upper limit of expected biogas availability for cogeneration because it has the maximum lime FOG quantity and dryer biogas use. Case B3 is representative of the lower limit of expected biogas availability for cogeneration because this scenario assumes both low WAS digestibility and high biogas utilization in the pelletizer. The rest of the cases cover various intermediate biogas production scenarios and were developed to provide insight and facilitate the decision process. The projected annual average and average weekly pattern of biogas balances for the above eight scenarios are presented in Table 1. It can be observed that in most of the cases the dryer operation is expected to be within the range of days, justifying the simplifying assumption that the available sludge and gas storage will be utilized so that the high biogas production period will be identical to the dryer operation period. The biogas balance results indicate that the amount of excess gas available for cogeneration is highly variable, depending on each scenario for FOG loading and biogas utilization in the dryer. For target scenario A4, an average of 361 scfm of excess biogas are projected to be available for cogeneration after the pelletizer biogas demand is met. In particular, 356 scfm would be available during pelletizer operation periods and 370 scfm would be available when the pelletizer would not be in operation. In the less likely lowest biogas production case (B3), there is not enough biogas generated to cover the energy requirements of both the pelletizer and cogeneration during days of pelletizer operation. For this case there is no biogas available for cogeneration during weekdays and there are 211 scfm when the pelletizer is not in operation, for an average of 72 scfm. When the energy in the excess biogas is considered, biogas availability for cogeneration would vary from 0 to 4,100 kw during dryer operation and from 2,400 to 4,200 kw during dryer downtime (typically during the weekends). BIOGAS QUALITY AND TREATMENT The decision on the level of biogas treatment needed for the implementation of cogeneration at SCBWRF was complicated by the fact that the digestion process would ultimately be changed to the three-phase digestion process, which is expected to alter the quality of the produced biogas. Based on the results from SCBWRF biogas sampling events performed in 2007 and 2010, during single-stage mesophilic operation of the SCBWRF ESDs, the hydrogen sulfide (H 2 S) concentrations were relatively high, in the range of 1,300 to 3,100 ppmv, while siloxane D4 concentrations were 0.19 to 0.2 ppmv and siloxane D5 concentrations were 1.6 to 17.1 ppmv. The measured biogas hydrogen sulfide levels are significantly above the 200 ppmv upper limit considered acceptable in the digester gas for use in lean-burn engines. However, it is expected that the new acid-phase reactors will help reduce sulfide concentration in the gas phase ESD biogas. In addition, the ESD biogas is expected to be controlled to values below 500 ppmv using ferric chloride addition to the ESDs, which is planned for sulfide and struvite control.

6 Table 1 - Projected Annual Average and Average Weekly Pattern of Biogas Case A1 A2 A3 A4 (Target) A5 B1 B2 B3 Percent of combined natural gas and biogas dryer fuel blend being biogas TWAS Ultimate VSR, % Number of 15,000 wet lb lime FOG truckloads per weekday WEDWRF Cake Addition Point Digesters Pelletizer Digesters Digesters Digesters Digesters Digesters Digesters Digesters FOG VS load, % of total VS load Methane phase (ESD) biogas produced (1), scfm Dryer in operation (weekdays), scfm Dryer not in operation (weekends), scfm % methane of biogas Days per week dryer operation Biogas used in dryer, scfm Dryer in operation, scfm Dryer not in operation, scfm Excess biogas available for cogeneration (2), scfm Dryer in operation, scfm Dryer not in operation, scfm Average power of excess biogas available for cogeneration (3), KW 3,000 2,600 3,500 4,100 2,100 1,700 2, Dryer in operation, KW 2,800 2,500 3,500 4,100 1, ,000 0 Dryer not in operation, KW 3,200 2,900 3,700 4,200 2,400 3,200 4,200 2,400 1 Only methane phase biogas to be utilized, the low quality acid phase biogas to be flared. 2 Assuming waste heat from cogeneration and dryer are sufficient to fully cover digestion heat demand and no flaring of methane phase biogas. 3 Converted by a cogeneration system to electrical power plus recovered heat plus heat losses.

7 The digestion process conversion to three-phase digestion with thermophilic ESD operation is expected to increase the siloxane content of the ESD biogas, mostly because of increased siloxane volatility as a result of the higher temperature of thermophilic digestion. Thermophilic digestion without an acid phase would significantly increase the level of siloxanes in the ESD biogas, as more siloxanes volatilize because of the higher thermophilic temperatures. This has been observed at several facilities and documented among others by Gary et al. (2001). Additional uncertainty in the prediction of the future siloxane content of ESD biogas exists because the feedstock blend will be altered after the phased digestion process conversion to include WEDWRF cake, lime FOG, and the uniform feeding of polymer FOG. Siloxanes content in wastewater may also increase in the future as a result of increased use of siloxane containing personal care products that are then washed into the wastewater system. Given the above uncertainty, it was decided to install complete biogas treatment train for all main contaminants, including biological sulfide removal, chilling to 40 F for moisture and partial siloxane removal, and activated carbon for final polishing including siloxane removal. SELECTION OF COGENERATION TECHNOLOGY The following alternatives for cogeneration were evaluated for SCBWRF. Internal combustion engines (ICE). Stirling engines. Fuel cells. Gas turbines. Microturbines. A preliminary analysis of the considered cogeneration technologies included the examination of their advantages and disadvantages and estimation of payback period and lifecycle financial benefit analyses. This initial comparison of the advantages of these cogeneration technologies indicated that the ICE technology has the most advantages for application at SCBWRF, as listed below (U.S. EPA, 2008). High electrical efficiency, in the order of percent, LHV basis (net output over lower heating value of input) with new lean-burn engines above 1 MW. Lower capital cost per kw. Can be brought on- and off-line quickly in response to real-time electrical power pricing. Routine maintenance is simple and can be performed by plant staff. Low gas inlet pressure required (3-5 psig). Most popular cogeneration alternative for biogas several available manufacturers. Tolerance of low siloxane levels, as opposed to turbines, microturbines, and fuel cells. As a result of these advantages, ICEs are the most frequently implemented cogeneration technology for water reclamation facilities in the range of mgd flow capacity, in which cases where the ICEs are generally rated at capacities of MW. Recent lean-burn ICE technology advancements allowed for significantly improved electrical efficiencies and reduced emissions.

8 The financial evaluations for the cogeneration technologies were based on the target scenario (Case A4). The Water Environment Research Foundation s (WERF) Life Cycle Assessment Manager for Energy Recovery (LCAMER) tool was used for the financial evaluation of the various technologies. The LCAMER model inputs for the ICE alternative were adjusted so that its predicted biogas quantities were in general agreement with the independently-determined biogas quantities for Case A4. The LCAMER tool, calibrated to SCBWRF conditions, was then applied to evaluate the remaining cogeneration technologies of Stirling engines, gas turbines, microturbines, and fuel cells. The financial comparison of the five cogeneration alternatives is presented in more detail by Surti et al. (2011) and is summarized in Table 2. The ICE technology s lifecycle financial benefit at about $5,500,000 was estimated to be more than twice the financial benefit from the secondbest technology for SCBWRF, the gas turbine technology (at about $2,400,000). In addition, ICE had the lowest total initial installation capital cost among all technologies. Based on this evaluation, the most suitable and promising cogeneration alternative for SCBWRF was determined to be ICE alternative and this one was further analyzed in more detail. Table 2 - Comparison of the Cogeneration Alternatives Using LCAMER Cogeneration system considered ICE GE Jenbacher J420 Stirling Stirling Biopower Gas Turbine Solar Saturn 20 Microturb ine Ingersoll- Rand MT250 Fuel Cell Fuel Cell Energy DFC1500 Number of cogen. units Net electrical capacity 1, , ,000 per unit, KW Avg. net capacity, KW 1, ,037 1,125 1,000 Electrical efficiency, % Total initial installation capital cost, $ $5,215,000 $5,608,000 $5,796,000 $5,675,000 $7,956,000 Cogeneration set service life, years Annual benefits from $1,076,000 $787,000 $835,400 $907,000 $805,700 power generation, $/year Total annual O&M cost, $357,000 $157,000 $285,000 $294,000 $413,000 $/year Net benefit, $/year $718,700 $630,000 $550,568 $611,900 $393, yr lifecycle net $5,480,000 $636,000 $2,400,000 $347,000 ($7,400,000) benefit (1), $/year Payback period (2), years Cost of Electricity: $0.092/kWh; Cost of Natural Gas: $0.0093/scf; Discount Rate (%): 3%; Capital and O&M cost includes the cost of biogas treatment for each cogeneration technology; 2 Including replacement capital cost

9 DETAILED EVALUATION OF ICE ALTERNATIVES Evaluation Approach There are several engine manufacturers of state-of-the-art high electrical efficiency lean-burn engines. These include Caterpillar, Cummins, GE Jenbacher, and Waukesha. These manufacturers produce ICE systems at manufacturer-specific power generation capacities, necessitating the evaluation to be performed for a specific manufacturer. The manufacturer selected for this evaluation was GE Jenbacher, which produces two ICE models that were considered suitable for implementation at SCBWRF. The evaluated ICE alternatives were as follows. Installation of a JMS420 ICE (1,426 kw rated capacity). Installation of a JMS320 plus a JMS 420 (2,485 kw combined rated capacity). Basic data for these two models are included in Table 3. In both cases, the ICEs are assumed to be operating at their rated capacity by supplementing the biogas with natural gas when needed. This would maximize the utilization of the installed capacity and allow the ICEs to operate at their maximum electrical efficiency. Table 3 - Basic Data for the 1,059 kw and 1,426 kw ICE Cogeneration Sets JMS320 JMS420 Gross output rated capacity, kw 1,059 1,426 Parasitic load (1), kw Total recovered heat available to heat the digesters, MBTU/hr Electrical efficiency at full load (2), % Thermal energy efficiency at capacity, % Net power plus thermal efficiency, % NOx, g/bhp-hr (3) NOx, tons/year@ 96% availability (4) Load consumed by the ICE generator set 2 Ratio of net electrical energy output at capacity to lower heat value (LHV) of input fuel 3 Versions with 0.6 g/bhp-hr NOx emissions are available at the same cost but have reduced electrical efficiency 4 Approximate value. Conceptual Project Cost Estimates Conceptual level cost estimates were generated and are summarized in Table 4. The biogas system considered is based on a design capacity of 370 scfm, which is the projected available

10 biogas for Case A4 during weekends. As such it remains the same for the either of the two cogeneneration systems in Table 4. Table 4 - Conceptual Level Cogeneration Project Cost Estimates Item Genset, accessories, and heat recovery equipment Electrical interconnection equipment and consolidation Biogas conditioning and treatment equipment JMS420 1,426 kw JMS320 Plus JMS 420 2,485 kw 1,150,000 2,100, , , , ,000 Total equipment cost $1,900,000 $3,000,000 Total project cost $5,200,000 $7,950,000 Comments Containerized Gensets with radiators, controls, silencers, gas trains, commissioning, etc. including exhaust heat recovery unit and accessories. Protective relays, transformers, switches, meters and distribution gear. Capacity of 370 scfm (target case A4). Adjusts pressure, temperature and moisture (40 F chiller) and removes particulates. Biological H 2 S removal system. Non-regenerative activated carbon. Conceptual O&M cost estimates Conceptual level O&M cost estimates were generated. The implemented cost factors and assumptions are presented in Tables 5 and 6, respectively. Table 5 - O&M cost factors for ICE Cogeneration Sets PCU daily oversight (based on continuous operation), FTE equivalent Genset system O&M expense Standard maintenance (oil changes) performed by plant personnel and major events and overhauls performed by contractor, $/hr Genset system O&M expense for consumables (e.g. oil and coolant), $/hr JMS320 1,059 kw JMS 420 1,426 kw

11 Table 6 Additional O&M Assumptions Item Value Comments ICE availability, percent 96 Power consumption for gas treatment, kw Power consumption for the heat loop pump, kw Hydrogen sulfide removal annual O&M cost Siloxane removal nonregenerative annual O&M cost 75 8 Considered an upper limit with good maintenance and gas treatment. Includes blower and chiller power. For 370 scfm system, scaled down linearly for lower scfm. $5,000 Based on biological system. $30,000 For biogas silicon being 5-10 ppmv. For 370 scfm system, scaled down linearly for lower scfm. Selection of the Number and Capacity of ICE Units The payback period and lifecycle financial benefits comparison of the two considered ICE alternatives (one vs. two ICE generation sets) was performed for the target case on available biogas (Case A4). This analysis is summarized in Table 7 for both the projected energy cost during the first year of cogeneration, as well as for a scenario anticipating a 25 percent escalation on the cost of energy. There is a tradeoff in the risk associated with the future cost of power and natural gas for the cases with one vs. two ICE systems. By operating solely on biogas, the one ICE system eliminates the dependency of the cogeneration on natural gas and the risk of its price fluctuations in the future. However, it generates less power and thus it results in a higher amount of purchased electrical energy for SCBWRF. Only about 40 percent of the total energy is generated by one ICE, compared to about 70 percent of the SCBWRF energy consumption generated by the two ICE alternatives. For the baseline unit cost of energy (9.2 cents/kwh and $0.9/therm), by operating solely on the biogas energy and thus not incurring any cost of purchasing natural gas for cogeneration, the alternative with one ICE is about 40 percent lower compared to the two ICE case (7.3 vs. 12 years). The payback period and lifecycle benefits of the alternatives with two ICEs are less attractive, due to the significant cost of supplemental natural gas (about $600,000 per year). Similar results are obtained if the unit cost of energy increases by 25 (9.2 cents/kwh and $0.9/therm), only in this case the payback period for the one ICE case is about 35 percent lower compared to the two ICE case (5.4 vs. 8.3 years). The reason for the lower percentage reduction in the payback period at the higher unit cost of energy is that the benefits for both cases increase significantly and reduce the importance of the cost to buy supplemental natural gas to use for cogeneration. The 20-year lifecycle benefits are significant in all four cases, ranging from $1.9 to $6.3 million. As a result of this analysis, and considering the uncertainty on the biogas quantity, PCU staff decided to implement the alternative with one ICE, with an installed rating in the order of 1,400 kw, with provisions for the potential installation of a second ICE sometime in the future. As a result, the remaining analysis focuses on the single ICE alternative.

12 Table 7 Comparison of Installing One or Two ICE Generation Sets. Case Baseline Electrical Energy Unit Cost JMS 420 JMS 320 JMS % Higher Electrical Energy Unit Cost JMS 420 JMS 320 JMS 420 ICE installed Available biogas case A4 A4 A4 A4 Electrical cost, cents/kwh Natural gas cost, $/therm Power rating, kw (1) 1,426 2,485 1,426 2,485 Avg. biogas available for cogeneration, scfm Avg. biogas used for cogeneration, scfm Percentage of available for cogeneration biogas (2) Energy NG supplement in cogen, % of total energy input to cogen (2) Generated electrical energy, kwh/year Percentage of the existing consumption at SCBWRF Annual value of produced electrical energy 11,698,000 20,351,000 11,698,000 20,351, $1,076,000 $1,872,000 $1,345,000 $2,340,000 Annual cost of NG supplement for cogeneration to operate at capacity - $611,000 - $764,000 Annual cogeneration O&M cost (3) $277,000 $501,000 $278,000 $502,000 Annual gas treatment O&M cost $86,000 $99,000 $99,000 $113,000 Annual net benefit from cogeneration $713,000 $661,000 $968,000 $960,000 Present worth of 20-year net benefits (4) $10,614,000 $9,838,000 $14,403,000 $14,282,000 Construction cost $5,215,000 $7,937,000 $5,215,000 $7,937,000 Payback period, years yr lifecycle benefits (4) $5,400,000 $1,900,000 $9,190,000 $6,350,000 1 Engines considered operating at capacity with supplemental NG when needed 2 The single JMS420 ICE case is sized at a fraction of the average available biogas to eliminate ICE operational dependency on NG throughout the year 3 Includes power for engine connection to the heat loop. 4 Based on discount rate (above inflation) of 3%.

13 Payback Analysis Based on 1,426 kw ICU The lifecycle benefits and payback periods were evaluated for the eight available biogas cases (A1 through A5 and B1 through B4) for one ICE installed (1,426 kw JMS420 for the purposes of these calculations) and based on the cost of a 370 scfm biogas treatment system, so that the maximum projected biogas case (Case A4) available for cogeneration biogas can be treated. The results are summarized in Table 8. The purpose of the analysis summarized in Table 8 is to estimate the risk in the payback period and lifecycle benefit if biogas is significantly reduced in the future for example due to a reduction in the available FOG quantities. The payback period and lifecycle benefit of the lower biogas cases are therefore penalized by the unutilized excess biogas treatment capacity and by the use of natural gas supplementation in the ICU so that it operates at its 1,426 kw capacity to fully utilize it. Cases A are considered more likely than Cases B, because Cases A are consistent with the SCBWRF design specifications. Among the A Cases, the ones with high TWAS digestibility have attractive payback periods and offer significant lifecycle net benefits ($3,200,000 to $5,400,000). The payback and lifecycle benefit of the lower biogas cases would approach those of Case A4 if each case is sized with its optimal (lower) biogas treatment and ICE capacities instead of the system being sized based on Case A4.The only exception corresponds to the lowest biogas case (Case B3), for which there no biogas surplus when the pelletizer is in operation, as shown in Table 1. The associated use of natural gas to run the ICE during such periods would therefore preclude the payback of case A4 from approaching that of Case A4. The target biogas case (A4), with three lime-fog truckloads per day, and alternative A3, with two lime-fog truckloads per day, have identical lifecycle benefits ($5,400,000) and payback period (7.3 years). This is because in both these scenarios there is excess biogas available beyond the ICE fuel demand. Based on Table 7, the 1,426 kw JMS420 ICE is sized conservatively to utilize only about 82 percent of the available for cogeneration biogas on the average, in order to minimize the probability of significant amount of supplemental natural gas needed in the future, given the long term uncertainty in the amount of available FOG and the natural gas price. Heat Balance A heat balance in the system was used to calculate the recovered heat from the 1,426 kw ICE as a fraction of the average heat demand of the digesters. The results are summarized in Table 9. The average heat recovered during the operation of the 1,426 kw ICE is projected to be sufficient to cover the average heat demand of the digesters. During periods of dryer operation, the waste heat from cogeneration will be added to the heat loop for heating the ESDs and for providing the relatively minimal heat demand of the polymer FOG storage tank. The acid reactors heat demand is expected to be covered by the waste heat from the dryer. During periods when the dryer is not operating, the heat recovered from the ICE will be used to heat the acid reactors, ESDs, and if not empty, the polymer-fog storage tank. The available thermal energy exceeds the heat demand of the system in all cases. As a result, the boiler will serve primarily as a backup heat source.

14 Table 8 Lifecycle Benefit and Payback Sensitivity Analysis on the Cost of Energy for the 1,426 kw (1) ICE Case A1 A2 A3 A4 (Target) A5 B1 B2 B3 Biogas in dryer fuel, % 35 (2) 35 (2) 35 (2) 35 (2) 35 (2) 90 (3) 90 (3) 85.4 (3) TWAS Ultimate VSR, % Number of lime FOG truckloads per weekday WEDWRF cake addition Digesters Pelletizer Digesters Annual value of produced electrical energy Annual cost of NG supplement to run cogen at capacity Annual value of not needing NG in boiler to heat ESDs Annual cogeneration O&M cost Annual gas treatment O&M cost Annual net benefit from cogeneration $1,076,000 Digesters Digesters Digesters Digesters Digesters $1,076,000 $1,076,000 $1,076,200 $1,076,000 $1,076,000 $1,076,000 $1,076,000 $159,000 $267, $422,000 $533,000 $240,000 $780, $74,000 $277,000 $74,000 $566,000 Present worth of 20-year net benefits (4) $8,420,000 $277,000 $277,000 $277,000 $277,000 $277,000 $277,000 $277,000 $66,000 $86,000 $86,000 $54,000 $46,000 $68,000 $28,000 $466,000 $713,000 $713,000 $323,000 $220,000 $491,000 $74,000 $6,939,000 $10,614,000 $10,614,000 $4,808,000 $3,272,000 $7,310,000 $1,098,000 Construction cost $5,215,000 $5,215,000 $5,215,000 $5,215,000 $5,215,000 $5,215,000 $5,215,000 $5,215,000 Payback period, years yr lifecycle net benefit (4) $3,210,000 $1,720,000 $5,400,000 $5,400,000 $(410,000) $(1,940,000) $2,090,000 $(4,120,000) 1 Covers roughly 40% of the present electrical demand of SCBWRF. 2 Scenarios A meet the dryer s design criteria. 3 Scenarios B are less likely to occur. 4 Based on discount rate (net interest rate above inflation) of 3%.

15 Table 9 Recovered Heat from the JMS420 ICE as a Fraction of the Average Heat Demand of the Digesters Case A1 A2 A3 A4 (Target) A5 B1 B2 B3 Biogas in dryer fuel, % TWAS Ultimate VSR, % Number of lime FOG truckloads per weekday WEDWRF cake addition (1) D P D D D D D D Average digesters heat demand (relative to 62% 55% 61% 61% 57% 62% 61% 57% cogenerated heat) Dryer operating (2) 47% 44% 47% 48% 44% 47% 48% 44% Dryer not in operation (3) 86% 82% 87% 88% 81% 86% 88% 81% 1 D: Digesters; P: Pelletizer 2 ESD only heat demand (Acid reactors heat demand covered by dryer waste heat, typically during weekdays) 3 Acid reactors plus ESD heat demand (typically during weekends) Sensitivity Analysis of Lifecycle Financial Benefit from Cogeneration The payback period and lifecycle benefits analysis was repeated for various scenarios of 25 percent unit price escalation for electrical power and/or natural gas. The results of this analysis are presented in Table 10. These results demonstrate the conservative nature of the selection of the 1,426 kw JMS420 ICE for scenarios A3 and A4, since these do not use supplemental natural gas and therefore their baseline payback does not increase with escalating natural gas prices, as opposed to the rest of the considered cases. The sensitivity of the financial benefits and payback period from Table 9 to the incremental addition of one, two and three truckloads of lime-fog per weekday is highlighted in the charts included in Figure 2 and Figure 3. From these, it becomes apparent that a 25 percent escalation in the price of electricity without an associated escalation in the price of natural gas, would make the use of natural gas supplementation to the ICE significantly more attractive and make the benefits of Case A1 approach the benefits of Cases A3 and A4. Conversely, a 25 percent escalation in the price of natural gas without an associated escalation in the price of electrical power would make the use of natural gas supplementation to the ICE significantly less attractive and reduce the lifecycle benefits of case A1 to roughly 50 percent of the benefits from cases A3 and A4. As a result, the lack of significant lime FOG quantities in the future is expected to reduce financial benefits by a large percentage only if there is a very large increase in the cost of natural gas relative to the cost of electricity. This could occur sporadically but is an unlikely long-term event, since natural gas prices influence electrical power costs

16 Table 10 Sensitivity Analysis of Lifecycle Benefit and Payback on the Cost of Energy for the JMS420 (1,426 kw) ICE Available Biogas Case A1 A2 A3 A4 (Target) A5 B1 B2 B3 Biogas in dryer fuel, % 35 (1) 35 (1) 35 (1) 35 (1) 35 (1) 90 (2) 90 (2) 85.4 (2) TWAS Ultimate VSR, % Number of 15,000 wet lb lime FOG truckloads per weekday WEDWRF Cake Addition Digesters Pelletizer Digesters Digesters Digesters Digesters Digesters Digesters Annual cost of NG supplement (3,4) $159,000 $267, $422,000 $533,000 $240,000 $780,000 Baseline (first year 2012) energy cost Power cost: $0.092/kWh; NG cost: $0.9/therm. Project cost = $5,215,000 Payback period, years year lifecycle benefits (5) $3,210,000 $1,720,000 $5,400,000 $5,400,000 $(410,000) $(1,940,000) $2,090,000 $(4,120,000) Power cost: $0.115/kWh (25% increase); NG cost: $1.13/therm (25% increase) Payback period, years year lifecycle benefits (5) $6,430,000 $4,570,000 $9,190,000 $9,190,000 $1,890,000 $(40,000) $5,040,000 $(2,780,000) Power cost: $0.115/kWh (25% increase); NG cost: $0.9/therm (no increase) Payback period, years year lifecycle benefits (5) $7,680,000 $6,670,000 $9,190,000 $9,190,000 $5,200,000 $4,150,000 $6,920,000 $2,700,000 Power cost: $0.092/kWh (no increase); NG cost: $1.13/therm (25% increase) Payback period, years year lifecycle net benefit (5) $2,610, $730,000 $5,400,000 $5,400,000 (1,980,000) $(3,930,000) $1,200,000 $(6,710,000) 1 Scenarios A meet the dryer s design criteria. 2 Scenarios B are less likely to occur. 3 Based on NG price of $0.9/therm. 4 Engine capacity is matched to cases A4 (target) and A3 based on operation without NG. 5 Based on discount rate (net interest rate above inflation) of 3%

17 Figure 2 Lifecycle Benefits Sensitivity Analysis 20 year Benefit $5,400,000 $5,400,000 $3,210,000 $9,190,000 $9,190,000 $6,430,000 $9,190,000 $9,190,000 $7,680,000 $5,400,000 $5,400,000 $2,610,000 Baseline: $0.092/kWh; $0.9/therm $0.115/kWh; $1.13/therm $0.115/kWh; $0.9/therm A4 (3 lime FOG truckloads per weekday) A3 (2 lime FOG truckloads per weekday) A1 (1 lime FOG truckload per weekday) $0.092/kWh; $1.13/therm Figure 3 Payback Period Sensitivity Analysis Payback Period Baseline: $0.092/kWh; $0.9/therm $0.115/kWh; $1.13/therm $0.115/kWh; $0.9/therm $0.092/kWh; $1.13/therm A4 (3 lime FOG truckloads per weekday) A3 (2 lime FOG truckloads per weekday) A1 (1 lime FOG truckload per weekday)

18 The effect of JMS420 1,426 kw ICE availability on payback and lifecycle benefits is summarized in Table 11. The target ICE availability of 96 percent was compared to a reduced availability of 93 percent. This reduction in ICE utilization results in a small increase of the payback period, from 7.3 to 7.5 years. Table 11 Effect of JMS420 1,426 kw ICE Availability on Payback and Lifecycle Benefits Target ICE Availability (96%) Reduced ICE Availability (93%) Available biogas case A4 (Target) A4 (Target) Electrical cost, cents/kwh Natural gas cost, $/therm ICE model JMS 420 JMS 420 Genset system O&M expense. Major events and overhauls outsourced to engine maintenance contractor, $/hr Power rating, kw (1) 1,426 kw 1,426 kw Annual value of produced electrical energy (1) $1,076,000 $1,042,600 Annual cost of NG supplement for cogeneration to operate at capacity - - Annual cogeneration O&M cost (2) $277,000 $268,000 Annual gas treatment O&M cost $86,000 $84,000 Annual net benefit from cogeneration $713,000 $691,000 Present worth of 20-year net benefits (3) $10,614,000 $10,277,000 Construction cost $5,215,000 $5,215,000 Payback period, years year lifecycle net benefit (3) $5,400,000 $5,060,000 1 Engines considered operating at capacity with supplemental NG when needed. 2 Includes power for engine connection to the heat loop. 3 Based on interest rate (above inflation) of 3%. The effect on the payback period and lifecycle benefits of the high efficiency vs. low emissions versions of the JMS420 (1,426 kw) ICE were evaluated. The low-emissions version reduces NOx emissions to 0.6 g/bhp-hr compared to 1.1 g/bhp-hr for the standard version, but has a lower electrical efficiency of 38.4 percent instead of 39.3 percent. Therefore, for the same net electrical power output, more fuel will be consumed by the low-nox JMS420 version. This is generally the case for high-efficiency and low-nox versions of ICE models from GE Jenbacher and other manufacturers, the exact tradeoff in efficiency loss being dependent on the considered ICE model.

19 The increased fuel demand would only result in additional cost if it results in an increased consumption of natural gas. For example, for the target Case A4 the low NOx version will likely consume more biogas without the need for additional natural gas. Therefore, the comparison of the high efficiency and low NOx versions of JMS420 was performed for Case A1 (based on receiving only one lime-fog truckload per day, about one third of the projected available), for which supplemental natural gas fuel is projected for the cogeneration system when operating at its capacity. The comparison is summarized in Table 12. For Case A1, the annual cost of natural gas supplement would increase by about $23,000 for the less-efficient low NOx version. The payback period would thus increase from 9.2 to 9.6 years, resulting in an associated reduction in the lifecycle benefits from $3,210,000 to $2,860,000. Table 12 Payback and Lifecycle Benefits Comparison of High Efficiency and Low Emissions Versions for Case A1 High Efficiency Version Low NOx Version Available biogas case (1) A1 A1 Electrical cost, cents/kwh Natural gas cost, $/therm ICE model JMS 420 JMS 420 Power rating, kw (2) 1,426 kw 1,426 kw Electrical efficiency, % NOx, g/bhp-hr Approx. NOx emissions at 96% availability, tons/year Annual value of produced electrical energy (2) $1,076,000 $1,076,000 Annual cost of NG supplement for cogeneration to operate at capacity $159,000 $182,000 Annual cogeneration O&M cost (3) $277,000 $277,000 Annual gas treatment O&M cost $64,000 $64,000 Annual net benefit from cogeneration $566,000 $543,000 Present worth of 20-year net benefits (4) $8,420,000 $8,075,000 Construction cost $5,215,000 $5,215,000 Payback period, years year lifecycle net benefit (4) $3,210,000 $2,860,000 1 Case A1 (1 lime-fog truckload per weekday). For 2 and 3 lime FOG truckloads per day, payback period is the same for both ICE versions. 2 Engines considered operating at capacity with supplemental NG; 96% ICE availability. 3 Includes power for engine connection to the heat loop. 4 Based on interest rate (above inflation) of 3%.

20 PCU will make the decision on the selection of a high-efficiency or low-nox ICE version during the next phase of the SCBWRF cogeneration project. CONCLUSIONS The economic feasibility of implementing combined heat and power generation at SCBWRF was analyzed in order to maximize lifecycle benefits and avoid energy wastage through biogas flaring. Analysis of various gas production scenarios considered the amount of available FOG as well as the weekly pattern of biogas production and consumption, presuming that the FOG would be fed to the digesters only during weekdays when the pelletizer would also be in operation. Sensitivity analysis of the payback period and lifecycle benefits on key technical and financial assumptions was performed. The main conclusions are as follows. Based on the expected high biogas production due to the codigestion of FOG and also due to a limit in the biogas utilization by the SCBWRF dryer/pelletizer, significant biogas quantities are expected to remain available after the pelletizer s biogas needs are met. When considering cogeneration coupled with the operation of a dryer, it is important to analyze separate periods when the dryer is or is not in operation to obtain a realistic estimate of the available biogas for cogeneration, since during dryer operation periods there may not be available biogas for cogeneration. ICE based cogeneration provided the maximum lifecycle benefit when compared to alternative technologies (Stirling engine, gas turbine, microturbine, or fuel cell technologies). The extreme case was considered, namely when there was no excess biogas projected to be available for cogeneration during dryer operation (Case B3), and involved the operation of the dryer on 85 percent biogas, with low TWAS digestibility (22.5 percent VSR), and low FOG quantities (14 percent of the total digesters VS load due to FOG). The selected ICE capacity was 1,426 kw rated electrical capacity, which allowed for operating on biogas and thus eliminating the dependency on the price of natural gas. Based on the target biogas scenario A4 (44 percent TWAS VSR and 26 percent of the total digesters VS load due to FOG), at $0.09/kWh, the payback period of a 1,400 kw (electrical) ICE cogeneration system and associated biogas treatment was estimated to be 7.3 years, with a projected 20-year lifecycle benefit of $5,400,000). At an electrical power cost of $0.115/kWh and NG cost of $0.9/therm, a cogeneration scheme that is sized conservatively to handle the maximum available biogas case would have a low payback period (5.4 to 9.8 years) for all available biogas cases. There is a tradeoff in the risk associated with the future cost of power and natural gas for the cases with one vs. two ICE systems. The one-ice system eliminates the dependency of the cogeneration on the risk of natural gas price fluctuations but generates less power and thus it results in a higher amount of purchased electrical energy for SCBWRF, increasing the risk from future electrical price uncertainty. ACKNOWLEDGEMENTS The authors would like to thank Pinellas County Utilities staff for their participation and help. GE Jenbacher staff provided valuable insights and information.

21 REFERENCES Gary, D.; G. Acosta; J. Kilgore; S. Min; Adams, G. (2001) Research Project to Remove Siloxanes from Digester Gas. Proceedings of the 74th Annual Water Environment Federation Technical Exposition and Conference [CD-ROM]; Atlanta, CA, Oct 13 17; Water Environment Federation: Alexandria, Virginia. Kabouris, J.C.; S.G.; Engelmann, M.; Dulaney, J.; Gillette, R.A.; Todd, A.C. (2008) Innovative FOG And Cake Digestion The Key To Biosolids Processing Energy and Cost Savings For Pinellas County. Proceedings of the 81st Annual Water Environment Federation Technical Exposition and Conference [CD-ROM]; Chicago, IL, Oct 18 22; Water Environment Federation: Alexandria, Virginia. Monteith et al. (2006) An Assessment Tool for Managing Cost-Effective Energy Recovery from Anaerobically Digested Wastewater Solids. WERF Project 01-CTS-18UP. Surti, J.R; Forbes, B.; Kabouris, J.; Filmore, L. (2011) Evaluation of the WERF Life Cycle Assessment Manager for Energy Recovery (LCAMER) Tool for Two Complex Facilities. Proceedings of the 25th Annual Water Environment Federation Residuals and Biosolids Conference [CD-ROM]; Sacramento, CA, May ; Water Environment Federation: Alexandria, Virginia. U.S. EPA (2008) Catalog of CHP Technologies. EPA Combined Heat and Power Partnership.