Analysis & Projections
Analysis of Heat Rate Improvement Potential at Coal-Fired Power Plants
Release date: May 19, 2015
Introduction
The thermal efficiency of electricity production is represented by the heat rate, which measures the amount of energy used to generate one kilowatthour of electricity.1 A generating unit with a lower, or more efficient, heat rate can generate the same quantity of electricity while consuming less fuel, compared with a unit with a higher heat rate. Lower fuel use per unit of electricity generated also reduces the corresponding emissions of pollutants such as sulfur dioxide (SO2), nitrogen oxide (NOX), mercury (Hg), and carbon dioxide (CO2). Consequently, improving heat rates at power plants can lower fuel costs and help achieve compliance with environmental regulations.
During the development of the Annual Energy Outlook 2015 (AEO2015), the U.S. Energy Information Administration (EIA) updated its modeling capability to include the ability to evaluate the potential for making heat rate improvements at existing coal-fired generators. The projections in the AEO2015 are produced by the National Energy Modeling System (NEMS), which is a modular system consisting of components to represent fuel supply, end-use consumption and conversion sectors, as well as modules for international and macroeconomic activities.
The Electricity Market Module (EMM) is the electricity supply component of the NEMS. The EMM performs three primary functions — capacity planning, fuel dispatching, and finance and pricing. Capacity planning decisions include building new plants to satisfy increases in demand and to replace retiring plants. Planning decisions also consider retrofits of existing capacity to install pollution control devices. The fuel dispatching function involves operating the available capacity to meet the demand for electricity. The finance and pricing function considers the investment costs associated with planning decisions and the operating costs from dispatching activities to develop delivered prices for electricity.
Heat rate improvement is another planning activity, as it considers the tradeoff between the investment expenditures and the savings in fuel and/or environmental compliance costs. Potential increases in efficiency can vary depending in part on the type of equipment installed at a generating plant. The EMM represents 32 configurations of existing coal-fired plants based on different combinations of particulate, sulfur dioxide (SO2), nitrogen oxide (NOX), mercury, and carbon emission controls (Table 1). These categories form the basis for evaluating the potential for heat rate improvements.
EIA entered into a contract with Leidos Corporation (Leidos) to develop a methodology to evaluate the potential for heat rate improvement at existing coal-fired generating plants. Leidos performed a statistical analysis of the heat rate characteristics of coal-fired generating units modeled by EIA in the EMM. Specifically, Leidos developed a predictive model for coal-fired electric generating unit heat rates as a function of various unit characteristics.2 Leidos employed statistical modeling techniques to create the predictive models.3
For the EMM plant types, the coal-fired generating units were categorized according to quartiles, based on observed4 versus predicted heat rates. Units in the first quartile (Q1), which perform better than predicted, were generally associated with the least potential for heat rate improvement. Units in the fourth quartile (Q4), representing the least efficient units relative to predicted values, were generally associated with the highest potential for heat rate improvement. Leidos developed a matrix of heat rate improvement options and associated costs, based on a literature review and the application of engineering judgment.
Little or no coal-fired capacity exists for the EMM plant types with mercury and carbon control configurations, therefore estimates were not developed for those plant types. These plant types were ultimately assigned the characteristics of the plants with the same combinations of particulate, SO2, and NOX controls. Plant types with relatively few observations were combined with other plant types having similar improvement profiles. As a result, 9 unique plant type combinations were developed for the purposes of the quartile analysis, and for each of these combinations Leidos created a minimum and a maximum potential for heat rate improvement along with the associated costs to achieve those improved efficiencies.5
Leidos used the minimum and maximum characteristics as a basis for developing estimates of mid-range cost and heat rate improvement potential. The mid-range estimates were used as the default values for the Annual Energy Outlook 2015 (AEO2015) (Table 2). Table 3 contains the minimum and maximum heat rate improvements and costs.
Additional details regarding the background and the analytical methodology are included in the consultant report prepared by Leidos Corporation (Appendix).
Plant Type | Particulate Controls | SO2 | NOX | Mercury Controls | Carbon Controls |
---|---|---|---|---|---|
B1 | BH | None | Any | None | None |
B2 | BH | None | Any | None | CCS |
B3 | BH | Wet | None | None | None |
B4 | BH | Wet | None | None | CCS |
B5 | BH | Wet | SCR | None | None |
B6 | BH | Wet | SCR | None | CCS |
B7 | BH | Dry | Any | None | None |
B8 | BH | Dry | Any | None | CCS |
C1 | CSE | None | Any | None | None |
C2 | CSE | None | Any | FF | None |
C3 | CSE | None | Any | FF | CCS |
C4 | CSE | Wet | None | None | None |
C5 | CSE | Wet | None | FF | None |
C6 | CSE | Wet | None | FF | CCS |
C7 | CSE | Wet | SCR | None | None |
C8 | CSE | Wet | SCR | FF | None |
C9 | CSE | Wet | SCR | FF | CCS |
CX | CSE | Dry | Any | None | None |
CY | CSE | Dry | Any | FF | None |
CZ | CSE | Dry | SCR | FF | CCS |
H1 | HSE/Oth | None | Any | None | None |
H2 | HSE/Oth | None | Any | FF | None |
H3 | HSE/Oth | None | Any | FF | CCS |
H4 | HSE/Oth | Wet | None | None | None |
H5 | HSE/Oth | Wet | None | FF | None |
H6 | HSE/Oth | Wet | None | FF | CCS |
H7 | HSE/Oth | Wet | SCR | None | None |
H8 | HSE/Oth | Wet | SCR | FF | None |
H9 | HSE/Oth | Wet | SCR | FF | CCS |
HA | HSE/Oth | Dry | Any | None | None |
HB | HSE/Oth | Dry | Any | FF | None |
HC | HSE/Oth | Dry | Any | FF | CCS |
Notes: Particulate Controls -- BH = baghouse, CSE = cold side electrostatic precipitator, HSE/Oth = hot side electrostatic precipitator/other/none. SO2 Controls -- wet = wet scrubber, dry = dry scrubber. NOX Controls -- SCR = selective catalytic reduction. Mercury Controls -- FF = fabric filter. Carbon Controls -- CCS = carbon capture and storage. Source: U.S. Energy Information Administration/Leidos Corporation |
Plant type and quartile combination | Count of total Units | Percentage HRI potential | Capital cost (million 2014 $/MW) | Average fixed O&M cost (2014 $/MW–yr) |
---|---|---|---|---|
B1-Q1 | 32 | (s) | 0.01 | 200 |
B1-Q2 | 15 | 0.8% | 0.10 | 2,000 |
B1-Q3 | 18 | 4% | 0.20 | 4,000 |
B1-Q4 | 20 | 6% | 0.90 | 20,000 |
B3-Q1 | 13 | (s) | 0.01 | 300 |
B3-Q2 | 24 | 0.7% | 0.05 | 1,000 |
B3-Q3 | 16 | 6% | 0.20 | 3,000 |
B3-Q4 | 15 | 9% | 0.60 | 10,000 |
B5C7-Q1 | 16 | (s) | (s) | 80 |
B5C7-Q2 | 42 | 0.8% | 0.03 | 700 |
B5C7H7-Q3 | 84 | 7% | 0.10 | 2,000 |
B5C7H7-Q4 | 59 | 10% | 0.20 | 4,000 |
B7-Q1 | 27 | (s) | (s) | 70 |
B7-Q2 | 25 | 0.8% | 0.04 | 800 |
B7-Q3Q4 | 30 | 7% | 0.30 | 5,000 |
C1H1-Q1 | 148 | (s) | 0.01 | 200 |
C1H1-Q2 | 117 | 0.8% | 0.10 | 2,000 |
C1H1-Q3 | 72 | 4% | 0.40 | 8,000 |
C1H1-Q4 | 110 | 7% | 1.00 | 30,000 |
C4-Q1 | 15 | (s) | (s) | 80 |
C4-Q2 | 27 | 0.8% | 0.04 | 900 |
C4-Q3 | 32 | 6% | 0.20 | 2,000 |
C4-Q4 | 39 | 10% | 0.30 | 5,000 |
CX-Q1Q2Q3Q4 | 15 | 7% | 0.20 | 4,000 |
H4-Q1Q2Q3 | 13 | 3% | 0.20 | 3,000 |
IG-Q1 | 3 | (s) | (s) | 60 |
Total set | 1,027 | 4% | 0.30 | 6,000 |
(s) = less than 0.05% for HRI potential or less than 0.005 million $/MW for capital cost. |
Plant type and quartile combination | Count of total units | Percentage HRI potential | Capital cost (million 2014 $/MW) |
---|---|---|---|
B1-Q1 | 32 | (s) | 0.007 |
B1-Q2 | 15 | 0.3% – 1.2% | 0.096 – 0.11 |
B1-Q3 | 18 | 2.1% – 6.4% | 0.20 – 0.26 |
B1-Q4 | 20 | 3.5% – 9.4% | 0.76 – 0.99 |
B3-Q1 | 13 | (s) | 0.010 |
B3-Q2 | 24 | 0.3% – 1.2% | 0.047 – 0.056 |
B3-Q3 | 16 | 3.1% – 8.2% | 0.19 – 0.30 |
B3-Q4 | 15 | 5.1% – 13% | 0.50 – 0.72 |
B5C7-Q1 | 16 | (s) | 0.003 |
B5C7-Q2 | 42 | 0.3% – 1.2% | 0.031 – 0.036 |
B5C7H7-Q3 | 84 | 3.6% – 9.5% | 0.11 – 0.16 |
B5C7H7-Q4 | 59 | 6.0% – 15% | 0.18 – 0.25 |
B7-Q1 | 27 | (s) | 0.002 |
B7-Q2 | 25 | 0.3% – 1.2% | 0.035 – 0.042 |
B7-Q3Q4 | 30 | 3.8% – 9.8% | 0.27 – 0.40 |
C1H1-Q1 | 148 | (s) | 0.006 |
C1H1-Q2 | 117 | 0.3% – 1.2% | 0.12 – 0.13 |
C1H1-Q3 | 72 | 2.0% – 6.0% | 0.36 – 0.49 |
C1H1-Q4 | 110 | 3.6% – 9.6% | 1.1 – 1.5 |
C4-Q1 | 15 | (s) | 0.002 |
C4-Q2 | 27 | 0.3% – 1.2% | 0.041 – 0.048 |
C4-Q3 | 32 | 3.5% – 9.1% | 0.13 – 0.20 |
C4-Q4 | 39 | 5.7% – 14% | 0.21 – 0.30 |
CX-Q1Q2Q3Q4 | 15 | 3.7% – 9.7% | 0.19 – 0.28 |
H4-Q1Q2Q3 | 13 | 1.9% – 5.1% | 0.14 – 0.21 |
IG-Q1 | 3 | (s) | 0.002 |
TOTAL SET | 1,027 | 2.0% — 5.3% | 0.24 — 0.32 |
(s) = less than 0.05% for HRI potential. |
Footnotes
1U.S. Energy Information Administration, Frequently Asked Questions, What is the efficiency of different types of power plants?, accessed January 31, 2015.
2The characteristics used to predict heat rate included attributes such as nameplate capacity, rank of coal used, NEMS plant type, flue-gas desulfurization status, and flue-gas particulate collector type.
3This included algorithmic evaluation of potential descriptive variables, and piecewise linear regression analysis. A decision tree created 7 sub-models describing inputs for the heat rate model for different unit categorizations.
4In this report, observed heat rates refer to the heat rates contained in EIA's EMM plant file.
5Leidos selected the plant type and quartile groupings such that each grouping contained at least 10 generating units, with the exception of the integrated gasification combined-cycle (IG) type, which has essentially no heat rate improvement potential. Some plant types and quartiles also had associated variable operation and maintenance (O&M) costs. The variable O&M costs were not incorporated into the NEMS EMM model at the time of this analysis. However, the impact of omitting variable O&M cost is expected to be small due to the relative magnitude of the capital and fixed O&M cost components.