Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
65 Chapter 6 Cost-Benefit Analysis The design and function of ferric oxide-sand filters evaluated in this study are largely the same as traditional sand filters designed for stormwater runoff treatment. As such, planning level life cycle costs of proposed ferric oxide-sand filters can be readily estimated using the life cycle costing tool provided as part of NCHRP 792. The primary differences between traditional sand filters and the ferric oxide-amended sand filters are performance and potential storage limitations (see Chapter 5) and the added cost of the ferric oxide media. Ferric oxide filter sizing will largely be based upon the volume of water desired treated with the primary limitation being the duration of ponding that can be provided before oxygen is reduced to zero and anaerobic conditions are established. To demonstrate the use of the NCHRP 792 life cycle cost tool (herein called the Tool) for ferric oxide-sand filters, the Highway 36/61 ferric oxide-sand filter was used as an example of how monitoring results supplied as part of this current study and the NCHRP 792 tool could have been used in the design phase of this project to calculate whole life cycle costs. Before the Tool can be employed, siting decisions and design detail will need to be defined to the extent required for the tool. The Tool can be used to calculate capital cost or if available project-specific costs can be used. Although it is expected that the Tool would be used in the early project planning phases, the Tool was applied to the Highway 36/61 site retrospectively as full design and post-construction data were available for this site, including: (1) completed designed drawings, (2) as-built construction costs, and (3) stormwater monitoring performance. Two designs were evaluated: (1) traditional sand filter, and (2) sand filter with ferric oxide media. The Tool provides whole life cycle annual costs per pound for several stormwater constituents including a few constituents relevant to this study: total suspended solids, total copper, total lead, and total zinc. It will be necessary to make final adjustments to the whole life cycle cost estimates if the project costs need to be expressed in terms of dissolved metals reduction. Also note that only copper, lead, and zinc are available for the Tool. For the example provided in this chapter, costs are presented in 2013 dollars consistent with the Tool and the construction bid tabulation costs for the Highway 36/61 ferric oxide-sand filter. The primary Tool inputs included (1) hydrologic parameters (Table 6-1), (2) runoff characteristics (Table 6-2), and (3) sand filter design parameters (Table 6-3). The hydrologic inputs are used to define the frequency and magnitude of runoff events for the identified site. If monitoring data are available for the site, the runoff coefficient can be adjusted to match the site data. It should be noted that the hydrology generated by the Tool is stochastic rather than deterministic. The stochastic approach calculates runoff volumes and event durations and water that cannot be infiltrated during storm duration will be bypassed. The deterministic approach uses a continuous flow record and may have different bypass volumes than the Tool depending upon individual storm event hydrograph peaks and the subsequent accumulation of water above the filter for each event. These differences are important to recognize as the procedure in Chapter 5 is deterministic and the life cycle cost estimate is stochastic. The runoff characteristics (Table 6-2) include monitoring data (e.g., chemical parameters) and the output of a calibrated hydrologic model (e.g., annual runoff volume). The ferric oxide-sand filter parameters in Table 6-3 are based upon Highway 36/61 ferric oxide-sand filter design drawings, field measurements at Highway 36/61, or Tool defaults.
66 Table 6-1 Highway 36/61 site and tributary watershed parameter inputs to the NCHRP 792 sand filter evaluation tool. Variables and Project Location Assumptions Value Units Source Location Maplewood, MN (1) Rain gauge ID 215435 (1) Elevation 872 feet (1) 85th Percentile, 24âHour Storm, Depth 0.80 inches (1) 95th Percentile, 24âHour Storm, Depth 1.33 inches (1) Average Annual Precipitation Depth 27.7 inches (1) Hydrologic Soil Group (HSG) Clay (D) none (2) Tributary Area (Total) 29.5 acres (3) Tributary Area Impervious 63 % (3) Tributary Area Impervious 18.6 acres (3) Calculated Runoff Coefficient 0.35 unitless (1) (1) NCHRP sand filter evaluation tool (2) Site soil borings (3) Design documents for the Highway 36/61 ferric oxideâsand filter Table 6-2 Highway 36/61 tributary runoff average and annual parameter inputs to the NCHRP 792 sand filter evaluation tool. Tributary Inflow Parameters Values1 Units Annual Runoff Volume 550,092 ft3 TSS 26 mg Lâ1 Total Zn 45.1 µg Lâ1 Total Pb 1.85 µg Lâ1 Total Cu 8.95 µg Lâ1 DOC 9.6 mg Lâ1 Average pH 6.8 s.u. (1) From 2018 monitoring, a user defined input. Runoff volume calculated using a hydrologic model for 2018. Table 6-3 Highway 36/61 ferric oxide-sand filter BMP design parameter inputs to the NCHRP 792 sand filter evaluation tool. Sand Filter Design Parameters Assumed Values Units Source Live Storage Volume 22,800 ft3 BMP Design Documents Max Ponding Depth 1.0 ft BMP Design Documents Media Thickness 1.2 ft BMP Design Documents Media Thickness (assumed by default) 2.0 (minimum) ft NCHRP Tool Media Infiltration Rate (k) 2.0 in hrâ1 2018 field monitoringÂ
67 Capital and operations and maintenance costs can be generated by the Tool or if site-specific capital and operation and maintenance costs are known they can be used as user defined inputs. As an example, project-specific capital costs (for the Highway 36/61 ferric oxide filer), project-specific operation and maintenance costs, Tool-provided capital costs, and Tool-provided operations and maintenance costs were generated and are provided in Table 6-4. The Tool-provided costs are a based upon Highway 36/61 ferric oxide-filter characteristics. Annualized project life cycle costs are estimated from capital costs, assumed BMP life, estimated operation and maintenance costs, and economic analysis parameters such as interest rates. For this analysis, the capital costs were user defined (e.g., based upon construction bids) while the costs calculated using the Tool are provided only for demonstration purposes. The user defined costs were developed based on detailed designs and unit prices that are benchmarked against 2013 construction prices for Highway 36/61 ferric oxide-sand filter. This estimate corresponds to a Class 5 or Class 4 cost estimate, characterized by up to 10 percent project definition (ASTM E 2516-11) and intended to serve as a reference in developing other planning level cost estimates using parametric methods and for benchmarking against other types of stormwater BMPs. Costs will change with individual project designs at other sites with differing soils, hydrology and constraints. The NCHRP 792 Tool was designed to incorporate these project-specific differences for sand filters and it is can be concluded that the Tool can be applied to ferric oxide-sand filters with minimal adjustment to both Tool inputs and outputs. The outcome of the whole life cycle cost analysis is provided in Table 6-5 . The comparative analysis was performed for a 22,800 square foot standard sand filter (capital cost $348,000) and a 22,800 square foot ferric oxide-sand filter ($416,000 capital cost, including ferric oxide media amendment) treating stormwater runoff from the Highway 36/61 tributary watershed. The surrogate standard sand filter results were modified to include the incremental cost increase for the iron enhanced media and the increased removal of total metals quantified by field monitoring. A range of BMP life (30 years and 50 years) and expected level of maintenance (high, medium, and low) were assumed to inform development of planning-oriented BMP cost ranges (maintenance designations identified in Taylor et. al., 2014). The cost ranges presented are intended to serve as an example of the type of lifecycle cost estimates that could be generated using the Tool for ferric oxide-sand filters. Interpretation of the results of this exercise suggests that ferric oxide-sand filters may provide incremental benefits of increased metals removal as a return on the incremental investment of the ferric oxide enhanced filtration media, indicated by higher removal efficiencies and possibly lower whole life cycle unit removal costs. Cost uncertainty in early stages of project planning and design is greater due to factors such as limited project definition, location, hydrology, uncertainty related to unit prices, uncertainty regarding design and analysis assumptions, limited on-site investigations, unforeseen constraints, and unforeseen constructability issues. In general, uncertainty will decrease as greater project definition is developed and more detailed information becomes available to reduce the uncertainty associated with these and other risk factors. Use and reference of the ferric oxide-sand filter cost information in this report should consider the sensitivity analysis (for maintenance intensity and BMP life) provided in this report when selecting an appropriate anticipated accuracy range for opinions of cost.
68 Table 6-4 Highway 36/61 life cycle cost analysis assumptions and inputs to the NCHRP 792 sand filter evaluation tool. Variables NCHRP Tool Inputs Value Units Source Cost Inputs Location Adjustment Factor Y 100 Minnesota factor per RS Means (105) Cost Year Y 2013 Sand filter tool default Discount Rate Y 3 % US Army Corps of Engineers Inflation Rate Y 1.5 % Minnesota Local Sales Tax Y 7.375 % Minnesota Cost Inputs (user defined) Capital Cost (sand filter construction) Y 244,000 $ User defined, 2013 bid tabulation Capital Cost (iron aggregate media) Y 68,000 $ User defined, 2013 bid tabulation Capital Cost (engineering and design, construction management) Y 104,000 $ User defined Capital Cost (all items) Y 416,000 $ User defined, 2013 Bid Tabulation & Owner Information Annual Operation & Maintenance N 21,627 $ yr -1 Owner Typical Cost (not used for this analysis, reference only) Cost Inputs (Tool Defaults) Capital Cost (all items, estimated using the NCHRP Tool) N 620,000 $ Sand filter tool default (not used for this analysis, reference only) Annual Operation & Maintenance (high effort expected)) Y 18,223 $ yr -1 Sand filter tool default cost database calculation Annual Operation & Maintenance (medium effort expected) Y 11,750 $ yr -1 Sand filter tool default cost database calculation Annual Operation & Maintenance (Low Effort Expected)) Y 10,314 $ yr -1 Sand filter tool default cost database calculation
69 Table 6-5 Highway 36/61 ferric oxide-sand whole life cycle cost estimate analysis. Constituent Condition1 Loading (lb yr-1) Standard Sand Filter Ferric Oxide Sand Filter Load Reduction (lb yr-1) Load Reduction (%) Annual Lifecycle Cost ($ yr-1) Whole Lifecycle Cost ($ lb-1) Load Reduction (lb yr-1) Load Reduction (%) Annual Lifecycle Cost ($ yr-1) Whole Lifecycle Cost ($ lb-1) TSS 30 / H 1051 731 70% $22,909 $31 893 85% $25,176 $28 30 / M 1051 731 70% $16,041 $21 893 85% $18,308 $20 30 / L 1051 731 70% $12,591 $17 893 85% $14,858 $17 50 / H 1051 731 70% $16,865 $23 893 85% $18,225 $20 50 / M 1051 731 70% $11,350 $15 893 85% $12,710 $14 50 / L 1051 731 70% $8,956 $12 893 85% $10,316 $12 Total Cu 30 / H 0.3621 0.0861 24% $22,909 $257,773 0.2498 69% $25,176 $100,765 30 / M 0.3621 0.0861 24% $16,041 $180,493 0.2498 69% $18,308 $73,276 30 / L 0.3621 0.0861 24% $12,591 $141,676 0.2498 69% $14,858 $59,468 50 / H 0.3621 0.0861 24% $16,865 $189,767 0.2498 69% $18,225 $72,944 50 / M 0.3621 0.0861 24% $11,350 $127,709 0.2498 69% $12,710 $50,871 50 / L 0.3621 0.0861 24% $8,956 $100,770 0.2498 69% $10,316 $41,289 Total Pb 30 / H 0.0749 0.0585 78% $22,909 $382,549 0.0689 92% $25,176 $365,357 30 / M 0.0749 0.0585 78% $16,041 $267,861 0.0689 92% $18,308 $265,688 30 / L 0.0749 0.0585 78% $12,591 $210255 0.0689 92% $14,858 $215,621 50 / H 0.0749 0.0585 78% $16,865 $281,625 0.0689 92% $18,225 $264,483 50 / M 0.0749 0.0585 78% $11,350 $189,527 0.0689 92% $12,710 $184,449 50 / L 0.0749 0.0585 78% $8,956 $149,548 0.0689 92% $10,316 $149,707 Total Zn 30 / H 1.8206 1.2436 68% $22,909 $17,973 1.6932 93% $25,176 $14,869 30 / M 1.8206 1.2436 68% $16,041 $12,584 1.6932 93% $18,308 $10,813 30 / L 1.8206 1.2436 68% $12,591 $9,878 1.6932 93% $14,858 $8,775 50 / H 1.8206 1.2436 68% $16,865 $13,231 1.6932 93% $18,225 $10,764 50 / M 1.8206 1.2436 68% $11,350 $8,904 1.6932 93% $12,710 $7,507 50 / L 1.8206 1.2436 68% $8,956 $7,026 1.6932 93% $10,316 $6,093 (1) Condition is the expected lifespan in years and maintenance effort as high (H), medium (M), or low (L). Estimated whole life cycle costs ($ lb-1) were observed to vary within a range, bounded on the low end by a 50-year life with low anticipated maintenance and on the high end by a 30-year life with high anticipated maintenance. Actual whole life cycle costs of constructed filters will vary, depending on the contributing watershed, runoff characteristics, actual hydrology, BMP sizing and level of maintenance required as well as the actual treatment cell lifespan. Lifespan may be affected by suspended solids loading as well as the rate at which ferric oxide binding sites are consumed. It is difficult to predict the binding capacity and lifespan of the ferric oxide media given several potential factors such as biding by other stormwater constituents (e.g., carbonates, chloride, and natural organic matter), formation of iron baring minerals, or leaching of iron from the filter.
