When does building an MBR make sense? How variations of local construction and operating cost parameters impact overall project economics Thor Young*, Sebastian Smoot*, Jeff Peeters**, Pierre Côté*** * GHD. 16701 Melford Blvd. Suite 330, Bowie, Maryland, 21401 USA. [email protected] ** GE Water & Process Technologies, Oakville, Ontario. *** COTE Membrane Separation Ltd.

Abstract This study details the expansion of a tool used to compare the capital and operating cost of membrane bioreactor (MBR) and conventional activated sludge (CAS) processes for a greenfield 5 MGD facility based on identical influent loading conditions for varying levels of effluent treatment objectives. Results from the original study (Young et al, 2012) are further developed to evaluate the impact of modifying key design assumptions and changing economic input variables such as the cost of labor, materials, equipment, land, and electricity. The sensitivity and scenario analyses presented in this study provide insight into which conditions tend to be most favorable for CAS or MBR facilities. Keywords: Membrane bioreactors (MBR), economic analysis, cost sensitivity, cost effectiveness.

INTRODUCTION Wastewater treatment facility owners, operators, and designers must consider many factors in selecting the treatment processes to use for a new facility including both economic and noneconomic considerations. Ultimately, however, many treatment process decisions come down to determining what is the most cost-effective approach for a particular site location, flow, climate, and effluent permit limitations. This economical evaluation should consider not only the initial capital cost of the facility, but the present value of the projected operation and maintenance (O&M) costs as well. Many of the capital and O&M costs of a wastewater treatment facility are sitespecific. Cost analyses should account for local market conditions when evaluating the costs of treatment alternatives. The use of membrane bioreactors (MBRs) for municipal wastewater treatment has expanded dramatically in the past decade, from an estimated 10 facilities over 1 mgd (3.8 ML/d) in 2001 to 166 facilities over 1 mgd (3.8 ML/d) in 2011 (Hirani et al, 2011). With advances in membrane technology, strategies to reduce operating costs, and increased membrane production, MBR treatment has become cost-competitive with conventional activated sludge (CAS) technology for situations which require lower effluent nutrient limits or water reuse (Young et al, 2012). This paper expands upon the prior work by examining how variations in key construction and operational cost parameters can impact the point at which MBR treatment is a more cost effective technology to use than CAS treatment to meet a particular set of effluent limits, temperature conditions, and flow variations. These parameters include: cost of purchasing land, equipment cost, materials cost, labor cost, electricity cost, membrane life, and discount rate. METHODOLOGY An Excel® based tool was developed previously (Young et al, 2012) to calculate and compare the capital, operating, and total lifecycle costs of different treatment strategies based on the same influent wastewater flow and load. The tool developed compares two biological process alternatives—one using a CAS process, and the other using an MBR process—for meeting a variety of different effluent requirements and design scenarios as described below.

Conceptual designs for both MBR and CAS alternatives were evaluated for seven different conditions, summarized in Table 1, which varied by effluent treatment objectives, primary clarification, minimum month wastewater temperature, and hydraulic peaking factors. Each condition has both a MBR and CAS alternative, resulting in a total of fourteen scenarios. For example, the MBR alternative for Scenario 1 is identified as “MBR-1” and the CAS alternative for Scenario 1 is identified as “CAS-1”. The basis of design for the liquid treatment process included 6 mm mechanical coarse screening, grit removal, activated sludge reactors, secondary clarifiers or membrane tanks for solids separation, and effluent UV disinfection. Two (2) mm fine screening was also included for all MBR alternatives following grit removal. Where target effluent limits required total nitrogen removal, a modified Ludzack-Ettinger (MLE) biological reactor configuration was used. The CAS scenarios with effluent limits for total phosphorus were designed with chemical addition of metal salts for phosphorus precipitation followed either by sand filters (for an effluent limit of TP < 0.2 mg/L) or a tertiary membrane filtration system (for TP < 0.1 mg/L). All treatment scenarios are designed for complete nitrification (effluent ammonia concentration less than 1 mg/L). Table 1. Summary of Conditions Evaluated for both MBR and CAS Alternatives Effluent mg/L Min Month Hourly Flow Scenario Description BOD/TSS/TN/TP Temp (°C) Peaking Factor 1 No TN or TP Limits 20/20 12 (54° F) 2 2 No TP Limit 20/20/10 12 2 3 Baseline ENR 10/10/10/0.2 12 2 4 Primary Clarifier 10/10/10/0.2 12 2 5 Warm Weather 10/10/10/0.2 25 (77° F) 2 6 High Peak Flow 10/10/10/0.2 12 4 7 Strict TP Limit 10/10/10/0.1 12 2 Conceptual designs were completed for greenfield wastewater treatment facilities using both MBR and CAS technology for the biological treatment. The systems were designed to treat 18,927 m³/d (5 mgd) of medium-strength wastewater. Concentrations were obtained from published typical values (Metcalf & Eddy, 2003). Process sizing and wastewater characterization were based on the midpoint of the published ranges of design criteria found in Metcalf & Eddy, Inc., 2003 and Water Environment Federation Manual of Practice No. 8, 2010. BioWin® process simulation software was used to size and configure the biological reactors as well as determine operating requirements such as aeration and chemical addition rates. The process was designed to provide a degree of redundancy typically provided for similar projects: • The membrane filtration system and sand filters were sized to pass the peak day flow with one unit out of service and the peak hour flow with all units in operation. • The fine screens and IPS pumps were designed with one fully-redundant spare unit. • The biological reactors were sized to meet the treatment objectives at the winter temperatures presented in Table 1 under average loading conditions with one unit out of service, and under maximum month loading conditions with all units in service. A hydraulic profile was created for all scenarios, and an intermediate pumping station (IPS) was determined to be necessary for facilities with sand filters in order to maintain the selected design criteria of an allowable 5 meter headloss through the entire facility on a flat site. Details of the design criteria are further discussed by Young et al, 2012.

Following completion of conceptual designs, capital costs were developed for the entire liquid treatment train for each alternative. The solids handling component of alternatives was assumed to be the same and therefore, costs for the solids process were not considered in the evaluation. Land requirements were estimated based on the size of the unit processes with appropriate allowance for setbacks. Annual operations and maintenance (O&M) costs of each alternative were estimated based on labor, projected electrical and chemical use, equipment maintenance, and replacement of diffusers, membranes, and UV lamps. The O&M labor cost for each scenario was developed based on the assumptions presented in Table 2. The baseline unit costs used in the analyses were based on the US national average costs provided by RS Means (2012). Process equipment costs were obtained from manufacturer quotes. Key unit costs used in the study are presented in Table 3. The labor, material, and equipment components of each unit cost were entered separately so that a sensitivity analysis could be performed on each component separately (i.e., raising the material cost of concrete while maintaining the labor and equipment at baseline values). Table 2. Baseline Operations Labor Cost Process Description Number of Staff, Conventional Activated Sludge Number of Staff, CAS with Tertiary Filters Number of Staff, CAS with Tertiary Membranes Number of Staff, Membrane Bioreactor Annual Cost per Employee (incl. overhead) Table 3. Baseline Unit Costs Value Parameter (US Units) Land $100,000 / acre Excavation $6.71/cy Backfill $8.04/cy Hauling $7.11/cy Concrete (Slab) $550/cy Concrete (Wall) $800/cy Handrail $76/lf Grating $44/sf Buildings $250/sf Paving $4/sf Equip. Installation 40% of equip. cost Membrane Installation 10% of membrane cost Electrical and Controls 25% of capital cost Yard Piping 10% of capital cost Site/Civil Works 5% of capital cost Electricity $0.10/kwh Alum Solution $0.10/lb Caustic Soda $2.16/gal Citric Acid $6.31/gal Methanol $1.38/gal Equip. Maintenance 2% of equip. cost/yr

Operators 5 6 5 4 $ 50,000

Value (SI Units) $247,105 / hectare $8.77/m3 $10.51/m3 $9.30/m3 $719/m3 $1,046/m3 $247/lf $472/m2 $2690/m2 $44/m2 40% of equip. cost 10% of membrane cost 25% of capital cost 10% of capital cost 5% of capital cost $0.10/kwh $0.22/kg $0.57/L $1.67/L $0.36/L 2% of equip. cost/yr

Mechanics 2 2 3 3 $ 70,000

Source Engineer estimate * RS Means (2012) RS Means (2012) RS Means (2012) RS Means (2012) RS Means (2012) RS Means (2012) RS Means (2012) RS Means (2012) RS Means (2012) Engineer estimate Engineer estimate Engineer estimate Engineer estimate ** Engineer estimate ** Engineer estimate * Supplier quote Supplier quote Supplier quote Supplier quote Engineer estimate

* An individual sensitivity analysis was performed for items marked with a single asterisk. ** Percentages for civil works and yard piping apply to CAS scenarios only. The percentages for civil works and yard piping for MBR scenarios were adjusted proportional to the relative area of their combined process tankage and structures as compared to that for the equivalent CAS scenario.

After development of these "baseline" cost models, some of the key design assumptions were revisited to assess their impact on the results. Following this analysis, the relative magnitude of each cost category (i.e., labor, materials, equipment, electricity, land, and chemicals) was assessed to determine which cost areas had the largest impact on overall project costs. For the initial screening, relative costs for Scenarios CAS-3 and MBR-3 were examined. A sensitivity analysis was then performed to assess the impact on the 20-year present-worth total lifecycle cost of varying the following values: cost of purchasing land, equipment costs, materials costs, labor costs, electricity costs. In the cost sensitivity analyses, construction equipment and process equipment costs were raised equally, as were construction labor and O&M labor. RESULTS AND DISCUSSION The results of the initial cost analyses are summarized in Figures 1, 2, and 3. Based on the "baseline" cost input values shown in Table 3, the 20-year present-worth total lifecycle costs (overall lifecycle costs) of MBR systems are less than those of CAS systems for plants designed for enhanced nutrient removal or water reuse. The higher O&M costs associated with MBR systems are offset by a lower capital cost for MBR systems compared to CAS systems. Raising the minimum design temperature or adding primary clarification before the bioreactors resulted in a slight improvement of the cost-competitiveness of the CAS alternative, but did not substantially alter the relative costs of the two systems. Where treatment requirements are less stringent, MBR systems have higher capital and O&M costs than CAS systems. Where the peaking factor is high, MBRs have a lower capital cost and a slightly higher overall lifecycle cost than CAS systems. MBRs were found to be most cost-competitive when compared to CAS followed by tertiary membrane filtration. Figure 1. Capital Cost Comparison by Treatment Process for All Scenarios Capital Cost, 2012 USD, Millions

$50 $45 $40 $35 $30 $25 $20 $15 $10 $5 $0

CAS

MBR

Scenario 1: No TN, TP Limit

Preliminary Treatment

Primary Clarifier

CAS

MBR

Scenario 2: No TP Limit

Biological Reactors

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

Scenario 3: Scenario 4: Scenario 5: Scenario 6: Scenario 7: Baseline ENR Primary Clarifier Warm Weather High Peak Flow Strict TP Limit

Chemical Feed

Secondary Clarifier

IPS

Sand Filters

Membrane System

UV

General

Figure 2. Present-Worth 20-Year O&M Cost Comparison for All Scenarios 20-Year Net Present Worth of O&M Costs, 2012 USD, Millions

$25

$20

$15

$10

$5

$0

CAS

MBR

Scenario 1: No TN, TP Limit

CAS

MBR

Scenario 2: No TP Limit

Labor

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

Scenario 3: Scenario 4: Scenario 5: Scenario 6: Scenario 7: Baseline ENR Primary Clarifier Warm Weather High Peak Flow Strict TP Limit

Power

Chemical

Equipment Maintenance

UV Lamp Replacement

Membrane Replacement

20-Year NPW, 2012 USD, Millions

Figure 3. Baseline Overall Lifecycle Cost Comparison for All Scenarios $80 $60 $40 $20 $0

CAS

MBR

Scenario 1: No TN, TP Limit

CAS

MBR

Scenario 2: No TP Limit

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

Scenario 3: Scenario 4: Scenario 5: Scenario 6: Scenario 7: Baseline ENR Primary Clarifier Warm Weather High Peak Flow Strict TP Limit Capital Costs

NPW of Lifecycle O&M Costs

Sensitivity of Varying Key Assumptions Prior to conducting the cost sensitivity analyses, three additional sets of results were generated to assess the impact of changing key assumptions from the original study: site topography, differences in O&M labor required for MBR and CAS processes, and the membrane replacement frequency. Site Topography. The original study assumed the need for an IPS for facilities with sand filters (Scenarios CAS-3, 4, 5, and 6) in order to keep structures from being excessively high above grade or deep on a flat site. If the site were sloped such that the hydraulic profile could match the site grading, the IPS could be avoided, thus reducing the cost of the CAS alternatives. The site topography does not affect scenarios with membrane filtration (All MBRs and CAS-7), as it was assumed that permeate pumps would be required to pass flow through the membranes. Figure 4

shows the overall lifecycle cost comparison for all scenarios with the capital, O&M, and land costs of the IPS removed. Removing the IPS makes the CAS-3, 4, and 5 scenarios nearly equal in cost to the respective MBR scenarios and results in CAS-6 being more cost-effective than MBR-6; the cost comparison for conditions with no TP limit and a TP limit of 0.1 mg/L remain unchanged.

20-Year NPW, 2012 USD, Millions

Figure 4. Overall Lifecycle Cost Comparison for All Scenarios (No IPS) $80 $60 $40 $20 $0

CAS

MBR

Scenario 1: No TN, TP Limit

CAS

MBR

Scenario 2: No TP Limit

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

Scenario 3: Scenario 4: Scenario 5: Scenario 6: Scenario 7: Baseline ENR Primary Clarifier Warm Weather High Peak Flow Strict TP Limit Capital Costs

NPW of Lifecycle O&M Costs

O&M Labor. The second part of this analysis was to examine the cost variation of operations and maintenance labor on the resulting overall lifecycle costs. The original assumption for the operations and labor for each alternative was shown in Table 2. In general, MBR facilities were assumed to require less operations labor (the process is typically highly automated) but more maintenance labor (MBR processes typically have more equipment and instrumentation). As CAS facilities added additional processes, such as sand filtration or tertiary membrane filters, the number of required operations staff was assumed to increase, based on discussions with superintendents of similarly-sized CAS and MBR facilities. However, even within the facilities surveyed, there was wide variation on the number of staff employed at each facility. Using the baseline assumptions, the cost results from each scenario were recalculated assuming the O&M labor cost for all of the scenarios was exactly the same (in this case, equal to that of MBR-3). Figure 5 shows the overall lifecycle cost comparison for all scenarios based on this assumption. Setting the CAS O&M labor costs equal to those of the MBR alternatives resulted in a negligible impact on the cost comparison.

20-Year NPW, 2012 USD, Millions

Figure 5. Overall Lifecycle Cost Comparison for All Scenarios (Equal O&M Labor) $80 $60 $40 $20 $0

CAS

MBR

Scenario 1: No TN, TP Limit

CAS

MBR

Scenario 2: No TP Limit

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

CAS

MBR

Scenario 3: Scenario 4: Scenario 5: Scenario 6: Scenario 7: Baseline ENR Primary Clarifier Warm Weather High Peak Flow Strict TP Limit Capital Costs

NPW of Lifecycle O&M Costs

As a further sensitivity analysis on the quantity of labor required, Figure 6 shows the overall lifecycle costs for scenarios CAS/MBR 1, 3, and 6 in which the labor required to run the MBR scenarios is held constant and the annual O&M labor costs for the CAS facilities is varied from 75%

to 125% of the baseline values presented in Table 2. As shown in the figure, adjusting the O&M labor for the CAS scenarios does not significantly affect the relative cost difference. Figure 6. Sensitivity Analysis of O&M Labor Costs for CAS Scenarios

Lifecycle Cost, Million 2012 Dollars

$100 $80 $60 $40 $20 $0 70%

80%

90% 100% 110% O&M Labor Cost Adjustment Factor for CAS

120%

130%

(O&M Labor for MBR maintained at constant level)

CAS-1

MBR-1

CAS-3

MBR-3

CAS-6

MBR-6

Membrane Replacement Frequency. The third part of the analysis was to examine the cost variation of varying the membrane replacement frequency. Over time, the membrane fibers experience a gradual loss of permeability and must be replaced periodically. The life expectancy of the membrane fibers is dependent on many factors, such as maintenance, influent loading, and water chemistry. Life expectancies for membranes given in the literature range from 6 years (Ayala, 2011) to 10 years or longer (Cote et al. 2012). As shown in Figure 7, the membrane replacement frequency does not have a significant impact on the overall lifecycle cost comparison between CAS and MBR at membrane life expectancies between 6 and 12 years. Figure 7. Sensitivity Analysis of Membrane Replacement Frequency

Lifecycle Cost, 2012 USD, Millions

$100 $80 $60 $40 $20 $0 0

2

CAS-1

4

6 8 10 12 14 Membrane Replacement Frequency (Years)

MBR-1

Sensitivity of Varying Key Unit Costs

CAS-3

MBR-3

16

CAS-6

18

MBR-6

20

Returning to the model results based on the original assumptions that were summarized in Figure 3, the relative magnitude of each cost for scenarios CAS-3 and MBR-3 are presented in Figure 8. Labor and materials account for a greater proportion of the overall lifecycle cost for the CAS-3 scenario as opposed to the MBR-3 scenario (approximately 60% vs. 50%, respectively). This finding suggests that MBR facilities are more cost-competitive in markets where the costs of labor and materials are relatively high. On the other hand, equipment and electricity account for a higher proportion of the overall lifecycle cost of the MBR alternative than the CAS alternative (approximately 40% vs. 30%, respectively). This finding suggests that CAS facilities are more costcompetitive in markets where the costs of equipment and electricity are relatively high. Sensitivity analyses were performed to evaluate the impact of adjusting the cost of labor, equipment, land, materials, and electricity; the results of each are discussed below.

OO&M Capital C OO&M Capital C

O MBR-3 C

O CAS-3 C

Figure 8. Relative Magnitude of Cost Categories for CAS-3 and MBR-3

$0

27%

20% 4%

15%

7%

3%

21% 15%

27% 6%

9%

$10

3%

21%

* Percentage values indicate portion of overall lifecycle cost associated with cost category. Values may not add to 100% due to rounding.

2%

15%

4% $20

$30

$40

$50

Capital Costs & Present-Worth of 20-year O&M Costs, Million USD (2012 Dollars)

Labor

Equipment

Land

Materials

Electricity

Chemicals

Labor. The cost of labor varies widely across the world. According to the US Bureau of Labor Statistics (2013), the hourly compensation cost in the manufacturing sector (which can be considered to be comparable to that of the construction sector) can range from over $60 per hour in Norway to $2 per hour in the Philippines; in the US, median manufacturing wages are approximately $36 per hour. Based on these values, a sensitivity analysis was conducted by adjusting the cost of labor from 5 to 180% of the baseline unit costs used in the initial evaluation. The overall lifecycle costs for scenarios CAS/MBR-1, 3, and 6 for various labor rates are shown in Figure 9. From this sensitivity analysis, one can make the following observations: • As the cost of labor increases beyond 130% of the baseline cost, the MBR alternative becomes more cost-competitive than CAS for the high peak flow condition (CAS/MBR-6). • In markets with low labor costs (less than 30% of the baseline cost), the cost of a CAS facility designed for 10 mg/L TN and 0.2 mg/L TP may be comparable to an MBR. • When nutrient removal is not required (CAS/MBR-1), the CAS alternative is more costcompetitive than an MBR even in markets with high labor costs.

Figure 9. Sensitivity Analysis of Labor Costs

Lifecycle Cost, 2012 USD, Millions

$100 $80

Brazil ($11) Mexico ($6) Phillippines ($2)

$60

Great Britain ($31) Singapore ($24) Israel ($20)

$40

Norway ($63) Switzerland ($58) Sweden ($50) Germany ($46) France ($40) United States ($36)

$20 $0 0%

25%

50%

75% 100% 125% Labor Cost Adjustment Factor

150%

175%

200%

Hourly wages (BLS, 2013) presented in parentheses for comparative purposes

CAS-1

MBR-1

CAS-3

MBR-3

CAS-6

MBR-6

Materials. As with labor, the cost of construction materials such as cement, steel, masonry, and aggregate varies widely across the world. As concrete is typically the most prevalent building material in wastewater treatment facilities, the price of concrete—which can range from less than $50 per ton in China to nearly $150 per ton in South America (ENR, 2012)—was used as a proxy for the overall cost of construction materials. The overall lifecycle costs for scenarios CAS/MBR-1, 3, and 6 for various labor rates are shown in Figure 10. The results of the sensitivity analysis of the materials cost are comparable to the results of the analysis based on the cost of labor: As material costs increase, the MBR alternatives become more cost-competitive. As shown in Figure 10, the cost of materials must fall to less than 25% of the baseline in order for the CAS alternative to become cost-competitive with MBR for Condition 3 (permit limit: 10 mg/L TN and 0.2 mg/L TP). Figure 10. Sensitivity Analysis of Material Costs

Lifecycle Cost, 2012 USD, Millions

$100 $80 $60 $40

Sao Paulo, Brazil ($146) New Delhi, India ($72) Doha, Qatar ($63) London, UK ($124) Shanghai, China ($49) Milan, Italy ($114) Sofia, Bulgaria ($39) Berlin, Germany ($98) Ankara, Turkey ($33) Los Angeles, USA ($84)

$20 $0 0%

25%

50%

75% 100% 125% Material Cost Adjustment Factor

150%

175%

200%

Cost of concrete per ton (ENR, 2012) presented in parentheses for comparative purposes

CAS-1

MBR-1

CAS-3

MBR-3

CAS-6

MBR-6

Equipment. In this analysis, "equipment" refers both to the equipment used during construction as well as the equipment that is permanently installed. While the cost of equipment can vary, it is not as site-specific as the cost of labor or construction materials, and the range in equipment cost is

estimated to deviate no more than 25% +/- of the baseline cost. This range is shaded in Figure 11. Only one scenario is sensitive to the variation in equipment cost within this range: CAS/MBR-6. Figure 11. Sensitivity Analysis of Equipment Costs

Lifecycle Cost, 2012 USD, Millions

$100 $80 $60 $40 $20 $0 0%

25%

CAS-1

50%

MBR-1

75% 100% 125% 150% Equipment Cost Adjustment Factor CAS-3

MBR-3

175%

CAS-6

200%

MBR-6

Electricity. The overall lifecycle costs for scenarios CAS/MBR-1, 3, and 6 for various electricity prices are shown in Figure 12. The price of electricity is highly dependent on many factors, including the type of fuel source, market conditions, government taxes and subsidies, regulations, and surcharges. As stated previously, MBRs are more energy-intensive than CAS processes; however, as shown in the figure, the price of electricity has a nearly insignificant effect on the relative cost difference between the MBR and CAS alternatives, even at very high or low prices. Figure 12. Sensitivity Analysis of Electricity Cost

Lifecycle Cost, 2012 USD, Millions

$100 $80 $60 $40

Mexico (¢12) Israel (¢10) Japan (¢18) New Zealand (¢7) Czech Republic (¢16) United States (¢7) Spain (¢15) Paraguay (¢6) Turkey (¢14) Kazakhstan (¢5) Great Britain (¢13)

$20 $0 $0.00

$0.05

$0.10 $0.15 $0.20 Cost of Electricity, USD per kWh

Slovak Republic (¢24)

Italy (¢28)

$0.25

$0.30

Electricity cost (¢/kWh) for industrial users (IEA, 2013) in parentheses for comparison

CAS-1

MBR-1

CAS-3

MBR-3

CAS-6

MBR-6

Land. The availability of land is often a critical factor when deciding between a CAS and an MBR process. MBR facilities are compact and occupy less area than their CAS counterparts. As shown in Figure 13, the cost of land only affects the cost comparison of CAS/MBR-6 (when the cost of land approaches $200,000 per acre). In order for MBR-1 to be cost-competitive with CAS-1, the cost of

land must approach $3M per acre. This is not surprising: as stated earlier, the cost of land was estimated to account for only 2-3% of the overall lifecycle cost of the entire facility. Figure 13. Sensitivity Analysis of Land Cost

Lifecycle Cost, 2012 USD, Millions

$100 $80 $60 $40 $20 $0 0

200,000

CAS-1

MBR-1

400,000 600,000 Cost of Land (USD per acre) CAS-3

MBR-3

800,000

CAS-6

1,000,000

MBR-6

CONCLUSION The results of the initial cost analyses are summarized in Figures 1, 2, and 3. Based on the "baseline" cost input values shown in Table 3, the 20-year present-worth total lifecycle costs (overall lifecycle costs) of MBR systems are less than those of CAS systems for plants designed for enhanced nutrient removal or water reuse. The results of this study indicate that the factor with the greatest influence on the relative costs between CAS and MBR facilities is the site topography: The overall lifecycle costs of CAS facilities are reduced by approximately 3% when the site is sloped such that an intermediate pumping station is not required. Eliminating the IPS reduces the difference in cost between the CAS and MBR alternatives for Conditions 3-6 but not for Condition 7 (TP limit of 0.1 mg/L). The second most influential factor was found to be the cost of construction and O&M labor. The overall lifecycle cost of CAS and MBR facilities were found to be nearly equal for Condition 3 (TN/TP limits of 10 and 0.2 mg/L, respectively) when the cost of labor was reduced to approximately 30% of the values originally used in the study. The cost of materials was found to have a similar, but lesser impact on the overall lifecycle cost comparison between CAS and MBR. While the overall lifecycle cost comparisons were heavily influenced by membrane replacement frequency and equipment cost, they were not significantly affected when variables were restricted to a range of expected values (the shaded areas in Figures 7 and 11). The costs of electricity and land and the assumed differences in the amount of O&M labor required for CAS and MBR facilities were not found to have a significant impact on the relative overall lifecycle cost comparison. In almost every sensitivity analysis, a "break-even point" was identified for CAS/MBR-6 (high peak flow condition); this is attributable to the fact that the overall lifecycle costs were nearly equal under the baseline assumptions and unit costs. No single adjustment of key design assumptions or key unit costs were found to result in MBR processes being cost-competitive with CAS processes

without tertiary filtration (Conditions 1 and 2; no TP limit). On the other hand, no single adjustment of assumptions or unit costs were found to result in CAS-7 (CAS followed by tertiary membranes) being cost-competitive with an MBR (Condition 7; TP limit of 0.1 mg/L). REFERENCES Ayala, D. F., Ferre, V., & Judd, S. J. (2011). "Membrane life estimation in full-scale immersed membrane bioreactors". Journal of Membrane Science, 378(1), 95-100. Cote, P.; Alam, Z.; Penny, J. (2012). “Hollow Fibre Membrane Life in Membrane Bioreactors”, Desalination 288:145-151. Cote, P.; Young, T.; Smoot, S. Peeters, J. (2013). "Energy Consumption of MBR for Municipal Wastewater: Treatment Current Situation and Potential." Proceedings of the WEF Energy and Water Conference; Nashville, TN. Engineering News Record (2012). "2012 4th Quarterly Cost Report". December 31, 2012. Metcalf & Eddy, Inc. (2003). Wastewater Engineering: Treatment & Reuse, 4th ed.; McGraw-Hill: New York, NY. Hirani, Z.; Oppenheimer, J.; DeCarolis, J.; Kiser, A.; Rittmann, B. (2010). "Membrane Bioreactor Effluent Water Quality and Technology - Organics, Nutrients and Microconstituents Removal". WEFTEC Proceedings. International Energy Agency (2012). "2012 Key World Energy Statistics". Available online at http://www.iea.org/publications/freepublications/publication/kwes.pdf. Accessed Aug 2013. RS Means Construction Publishers & Consultants. (2012). Facilities Construction Cost Data, 28th ed. Reed Construction Data, Inc: Norcross, GA. United States Bureau of Labor Statistics (2013). "International Comparisons of Hourly Compensation Costs in Manufacturing, 2012". International Labor Comparisons. Water Environment Federation. (2010). Design of Municipal Wastewater Treatment Plants. WEF Manual of Practice No. 8 (5th Ed). Water Environment Federation, Alexandria, VA. Young, T.; Muftugil, M.; Smoot, S.; Peeters, J. (2012), “Capital and Operating Cost Evaluation of CAS vs. MBR Treatment”. WEFTEC Proceedings.