By: Bill Hoskins, P.E., C.C.E., Washington Group International
Numerous FGD systems are being planned in order to comply with continually more stringent power plant emissions regulations. Plant operators can choose between wet and dry flue gas desulfurization technologies. Although wet FGD systems are more common, dry FGD using “dry” absorbers and fabric filter (FF) systems are being considered more often because of simplified operation, combined SO2 and particulate control, and dry waste product. In addition, dry FGD systems are particularly well suited to plants burning low and medium sulfur coal.
This article evaluates the capital costs and technical features of eight dry FGD installations in the U.S. These systems entered operation within the past 10 years and range in size from about 150 MW to 500 MW. The installations are on coal-fired units and include lime spray dryer absorbers (SDA) and either reverse gas or pulsejet fabric filters (FF) — seven of the eight installations have reverse gas FFs.
The available cost, performance, and technical data in this report were gathered from published sources including:
- U.S. Energy Information Administration (EIA) tabulations, as reported by utilities on EIA Form 767;1
- Published technical papers from POWER-GEN International and other conferences;
- Publicly available material from state public utility commissions and state environmental agencies; 2 and/or
- Telephone contacts with personnel at state public utility commissions and state environmental agencies.
A significant portion of the technical information and some of the cost data were obtained from EIA Form 767. These data were utilized in this analysis or in some cases modified if two or more independent sources provided differing data.
Although these are actual installations and the information comes from published sources, the units are not identified by name because of sensitivity to the competitiveness of the “deregulated” electric market.
New vs. Retrofit
The dry FGD systems evaluated include new, retrofit, and brownfield installations (brownfield is defined as new pulverized coal boiler, steam turbine and associated systems that are added to an existing plant site). As expected, for systems of similar size, new installations are not as costly as retrofit or brownfield installations. Retrofit installations incur additional costs because the dry FGD systems must be fit within existing site space and must interface with existing plant systems and/or structures. Compared to new installations, many retrofit installations have limited site space and physical interferences, which results in:
- Restricted construction access,
- Less efficient construction sequence,*
- Longer, higher, or unconventional duct runs, and
- Equipment layout that is more complicated and more costly to construct than at a new plant.
The analyses in this article provide summary-level subsystem size information: identification of the number of trains of the SDA systems, lime preparation systems, and fans; FF systems; and total constructed cost. It was not possible in all cases to obtain separate costs for the balance-of-plant (BOP) systems, consisting of lime receiving and handling, waste handling, ductwork, and other items associated with the SDA. Therefore, each SDA system is represented by the one cost, which is the combined cost of SDA(s) and BOP equipment.
The tables and other material in this article broadly represent “similar” system scope and show corresponding respective costs for the SDA + BOP and FF systems. However, the competitive nature of the power generation industry made it impossible to obtain detailed listings identifying all components included or excluded from the individual systems. Therefore, the costs of SDA + BOP and FF systems herein should be considered in terms of broad indicators of benchmark costs and/or as providing relative comparison of retrofit costs to new costs. Even with these limitations, however, the results show that economies of scale apply to retrofit systems.
The initial analysis identified 13 coal-fired units with SDA/FF systems that started operation during or after 1994. However, technical and cost information sufficient to enable meaningful comparison was only available for dry FGD/FF systems on eight units. The unit names are indicated by capital letters and not by actual unit or plant name to maintain anonymity as previously indicated. The unit letters were assigned to the initial list of 13 units and are not consecutive because five units were eliminated during the evaluation process.
Table 1 shows that the installed costs of SDA + BOP retrofits for Units A, B, and F range from $140/kW to $230/kW (based on 2002 dollars, for SDA + BOP only — excluding FF). Table 1 also presents the total installed costs of SDA + BOP + FF systems escalated to the common date (2002 dollars). The total costs for the new, retrofit, and brownfield installations at units C, D, E, J, and K range from $150/kW to $320/kW. The costs are escalated to the common date with published technology-specific cost escalation factors applicable to the respective systems.
Figure 1 provides a graphical comparison of the installed costs of SDA + BOP + FF systems. This figure illustrates that the total installed cost of retrofit and brownfield systems at Units C, E, and K are higher than the new systems at Units D and J. The higher costs of these retrofit units are influenced by the constrained site space, limited construction access, and “non-optimum” layout of major systems, equipment, and ductwork.
To address the potential quantitative impact of a dry FGD retrofit, the costs of two similar size units from Figure 1 were compared. The ratio of Unit E (retrofit) to Unit D (new) is about 1.37 (cost of Unit E divided by cost of Unit D). However, this should only be considered as a broad indicator of possible retrofit cost increases because the units are located in different states and no adjustments have been made for differences in labor and productivity or train size. Even so, the comparison indicates the trend in cost that would be expected between retrofit systems and new systems of similar size.
Agreement with EPRI Cost Model
Figure 2 compares the installed retrofit costs of three existing SDA + BOP systems to the installed retrofit costs estimated by EPRI’s FGDCOST program.3 The FGDCOST program was used to estimate total installed costs of SDA + BOP systems for hypothetical unit sizes ranging from about 100 MW to about 500 MW. The cost estimates from FGDCOST were based on a retrofit factor of 1.3. The cost of each hypothetical new system was multiplied by a factor of 1.3 to obtain a cost representative of a retrofit system.
The retrofit cost reflects the increase resulting from site space limitations, complications in interfacing with existing plant systems and/or structures, and modifications in construction sequence or methods necessary to accommodate the retrofit. The factor of 1.3 is representative of a moderate to moderately difficult retrofit, which is one of the “standard” retrofit options available in the FGDCOST program.
It is important to point out that Figure 2 does NOT include the cost for the fabric filter system. Also, the dry FGD systems for units C, D, E, J, and K are not included in Figure 2 because of the lack of separate retrofit costs for SDA + BOP systems or because the systems were installed on brownfield or new units.
From Figure 2, it is apparent that the costs estimated by the FGDCOST program for SDA + BOP systems are generally consistent with the costs of the existing SDA + BOP systems (for the units analyzed in this article). The figure shows that the costs for the systems on Units A, B, and F are 14% less, 18% more, and 12% more than the respective costs estimated by FGDCOST. The differences in retrofits for other emission control systems, such as selective catalytic reduction, has shown that constructed costs can be as much as 30% lower or 40% higher than a curve fit of more than 20 units.4 Therefore, the SDA + BOP systems on Units A, B, and F are within the expected range of variance from the curve predicted by the FGDCOST program. In addition, it appears that the FGDCOST program is a useful tool for estimating costs of SDA + BOP retrofits for conceptual or scoping studies, especially when the potential cost variance is kept in mind.
The information from Figure 2 also shows that the difference in cost between the retrofit systems on Units A and B is about 22% even though the units/systems are similar in size. In addition, Table 1 shows that the respective number of SDAs and FFs for these two units is the same and the respective size is similar. Information and plan views from various published references indicate that the available space at Unit B is more constrained than the available space at Unit A. This illustrates the impact that constrained site space can have on retrofit costs.
Data on bulk material cost breakdowns, total erection labor hours, or erection labor rates could not be obtained for the FGD cost analysis performed in this study. Therefore, it was not possible to adjust construction costs or bulk material costs to a common location allowing comparison on a more common footing. If this information was available, the difference in costs of SDA + BOP retrofits of similar size could be narrowed. On the other hand, such adjustments might not make a significant impact on the differences. In general, the custom nature of retrofits, differences in site space, and location/capacity of existing equipment have a greater impact on system capital costs than on erection costs.
Therefore, reported costs of dry FGD retrofits of similar size should not be characterized by single rule-of-thumb costs or even narrow cost ranges. Rather, comparisons in this article have shown that retrofit costs should be considered in terms of broad cost indicators or as providing “representative” cost boundaries for SDA + BOP systems. In addition, although the number of units in this evaluation is limited, it appears that retrofit costs of dry FGD systems can be 20% to 40% higher than a new unit of similar size. This range of cost increase is a good starting point that is useful in defining cost increases of retrofits of dry FGD systems, especially for studies at the conceptual or planning stage.
1. Raw data files developed by EIA from EIA Form 767 (available from EIA web site).
2. Various documents, articles, and information from POWER-GEN International, the Mega Symposium, Public Utility Commissions, and State Environmental Control agencies that are not identified by specific reference or unit name because of sensitivity to the competitiveness of the “deregulated” electric market.
3. FGDCOST, Version 5.00, EPRI Product ID Number 1006621, November 2002.
4. Hoskins, W., “Comparison of Costs and Plant-Specific Factors Influencing Costs of SCR Retrofits on Coal-Fired Power Plants,” Power Engineering, May 2003.
Bill Hoskins, P.E., is a supervising process engineer at Washington Group International (Integrated Engineering, Construction and Management Solutions). He is responsible for scoping and feasibility studies for the power industry. Hoskins is a Certified Cost Engineer (CCE) and holds a bachelor’s degree in chemical engineering from New Mexico State University.