Conservation Supply Curves for Industrial Energy Use
Relatively little analysis has been undertaken to address systematically the costs of achieving energy savings and consequent CO2 emission reductions for the industrial sector. This is due in part to the extraordinary heterogeneity of this sector (e.g., in contrast to the buildings or transportation sector) and to the often proprietary nature of industrial process technology. Thus, although technical aspects of industrial energy use have been studied extensively, economic assessments remain largely lacking at this time, especially with regard to fundamental process design changes, which offer some of the potentially most significant energy reduction measures.
The economic analyses available focus primarily on the use of electricity in selected manufacturing processes. A key assumption in such analyses is the rate of return, or discount rate, needed to induce capital investments in energy conservation. Empirically, the discount rates required for the industrial sector appear to be on the order of 30 to 50 percent, equivalent to payback periods of only 2 to 3 years (Ross, 1990b). In some cases, the desired rates of return may be as high as 100 percent (Ayres, 1990). Although financial decision-making criteria in industry are not very well understood, U.S. manufacturers appear to be more reluctant to invest in energy-saving equipment than would be suggested by standard financial analyses using more typical rates of return (e.g., 10 to 20 percent). Part of this behavior stems from the fact that energy costs for most industries represent only a small portion of overall expenses. Hence, the reliability of energy supplies is often more critical than energy costs. For the United States, there is little empirical basis to assess investment behavior for substantially
higher energy prices than those now prevalent (e.g., electricity prices 2 to 3 times higher).
An estimate of the electricity conservation potential of the overall U.S. manufacturing sector has been developed by Ross (1990a), based on an aggregation of results for aluminum production, fabrication/assembly processes, and other selected process industries, and assuming a capital recovery rate of 33 percent. Figures D.1 to D.3 show the electricity conservation supply curves first developed for individual manufacturing industries. The curve for fabrication and assembly (Figure D.1), developed from studies of the automobile industry, was assumed to apply to a number of other industry groups as well. The results from Figures D.1 to D.3 were then combined by using energy weights based on 1985 electricity consumption (i.e., 1260, 190, and 768 × 1012 Btu for process industries, aluminum, and fabrication/assembly, respectively (Ross, 1990b)). This yielded the overall manufacturing sector curve shown in Figure D.4.
The curve in Figure D.4 can also be transformed into a cost-effectiveness curve for CO2 reductions similar to those presented in Chapter 21 for residential and commercial buildings. The electricity savings percentages on the x-axis are first converted to kilowatt-hours by employing the total 1985 manufacturing electricity use of 653 billion kilowatt-hours (BkWh) (which was the basis for the estimates of Ross). Current (1989) electricity use is
roughly comparable in magnitude. Then, a utility fuel mix must be assumed to convert kilowatt-hours (on both axes) to equivalent CO2 emissions. Figure D.5 shows the results based on the current U.S. average fuel mix and carbon content for electricity production (equivalent to 0.7 gCO2per kilowatt-hour consumed, as derived in Chapter 21). By using Ross's original curve for a 33 percent capital recovery factor (which corresponds to a discount rate of 30 percent over a 10-year project life), curves for real discount rates of 3, 6, and 10 percent were also developed, based on a 10-yearproject life. The capital recovery factors (CRFs) for these three discount rates are 0.117, 0.136, and 0.163, as found from the expression
where d is the discount rate and n is the project life. Because annual operating and maintenance cost savings are negligible in Ross's cost curve, the results for a 33 percent CRF can be adjusted to other values by simple proportion:
where p1 and p2 are the equivalent electricity prices of the annualized capital investment.
As discussed earlier for the buildings sector, the discount rates of 3, 6, and 10 percent do not reflect actual consumer practice. Rather, they are intended to indicate what might be possible if government policies could induce action at these effective rates of return. Thus the maximum electricity savings of 30 percent (200 BkWh) reduces CO2 emissions by about 140 megatons (Mt) at an average cost of about $175/t CO2 at a 30 percent discount rate, $85/t at a 10 percent discount rate, and about $60/t at 3 percent.1
Offsetting this cost is the savings in electricity use. The value of this savings depends on the price of electricity. By following the method elaborated in Chapter 21, two electricity prices are used to estimate a net cost of CO2 reduction. For the 30 percent discount rate, reflecting actual industrial practice, the current average price of electricity for the industrial sector (4.8 cents/kWh) is used to reflect typical savings to industry. For the lower discount rates, reflecting a societal perspective, the average U.S. electricity price of 6.4 cents/kWh is used to estimate savings. As discussed in Chapters 21 and 22, the current average electricity prices are implicitly taken to be a long-run marginal cost of supply. To the extent that actual marginal costs are higher in the future, the current estimates of net cost are conservative.
For the maximum electricity savings of 200 BkWh, Figure D.5 shows a net negative cost for the 3, 6, and 10 percent discount rates of -$30, -$20,
and -$7/t CO2, respectively. For example, at a 6 percent rate of return, the cost of conservation is equivalent to 5 cents/kWh, but the value of electricity saved is 6.4 cents/kWh, leading to a net savings of -1.4 cents/kWh. Dividing by the fuel mix factor of 0.7 kg CO2/kWh and then adjusting units yield the net savings of -$20/t CO2. Similarly, a net positive cost of $106/t is found for the 30 percent discount rate, based on an actual electricity cost of 4.8 cents/kWh for the industrial sector. Note that significant uncertainty still surrounds all these estimates and that a more refined analysis would also have to consider the fuel mix on a regional basis, as noted in Chapter 22.
The electricity savings and costs derived from Ross (1990a) are comparable to those reported by the Electric Power Research Institute (EPRI) in a study of the maximum technical potential for electricity savings (Electric Power Research Institute, 1990). Figure D.6 shows the composite conservation supply curve developed for the residential, commercial, and industrial sectors. For industry, a maximum electrical savings of about 277 BkWh was estimated for the year 2000, corresponding to 24 to 38 percent of projected industrial electricity use, depending on the assumed growth rate. The technologies considered in the EPRI analysis were largely the same as those considered by Ross: adjustable speed drives for electric motors, high-efficiency motors, waste heat recovery including recouperators, diaphragm and membrane cells in chlor-alkali production, more efficient electrolytic cells in aluminum reduction, and more efficient lighting technologies (Barakat and Chamberlin, Inc., 1990).
Figure D.6 shows the average cost of the energy conservation measures to be about 3 cents/kWh for a 5 percent real rate of return. This is similar to the costs shown in Figure D.5 at a comparable rate of return. Again, however, it must be stressed that significant barriers remain to actually achieving the magnitude of savings suggested by Figures D.5 and D.6.
Rough estimates of the cost and CO2 reduction potential from increased use of co-generation in the industrial sector are derived based on the assumptions in Table D.1. This approach takes the societal perspective discussed earlier, and reflected in the Chapter 29 results using a 6 percent real discount rate. Thus the fuel savings and investment cost of co-generation are evaluated relative to a reference case where steam and electricity are generated separately using natural gas.
Using the data in Table D.1, the annualized capital cost is $87.2/kW-yr and the levelized annual fuel savings is $140.9/kW-yr, yielding a net cost of -$53.7/kW-yr. With no fuel cost escalation the net cost is -$11.4/kW-yr. The corresponding range of CO2 reduction costs (based on the carbon content
of natural gas) is -$30 to -$6/t CO2. Based on an incremental co-generation capacity of 25,000 MW, the total fuel energy savings would be approximately 0.82 quads. For natural gas the corresponding overallCO2reduction would be 45 Mt. To the extent that oil rather than gas is the fuel backed out in some co-generation applications, the corresponding CO2reductionwould be somewhat greater.
TABLE D.1 Assumptions for Co-generation Analysisa
Other Energy Savings
Industrial fuel savings from improvements to current process operations are estimated using the results of three case studies of energy-intensive industries (Ross, 1987; Larsen, 1990; K. C. Nelson, Dow Chemical USA, Plaquemine, Louisiana, personal communication, January 1991). These studies are used to estimate the net cost of achieving different levels of total energy savings. In the absence of more comprehensive data for the industrial sector, rough estimates of the overall CO2 reduction potential for the industrial sector are obtained by applying these results to current (1989) CO2 emission.
1. Steel Mill (Ross, 1987). From Figure 22.9a (Chapter 22), the most costly project has an incremental cost of $14/t-yr and an energy savings of 3 percent. The reported total specific energy for this plant is 31 MBtu/t; so the energy savings is 0.93 MBtu/t. Assuming a project life of 10 years and a discount rate of 6 percent (CRF = 0.136), the project investment cost is $2.05/MBtu saved. Escalating this cost from 1982 dollars to the present gives a cost of approximately $2.60/MBtu for a 25 percent energy savings. Similarly, using Figure 22.9b for a 15 percent savings, the cost is approximately $1.00/MBtu.
2. Petroleum Refinery (Larsen, 1990). Use Figure 22.9a and perform the same calculations as above. The reported total specific energy for this plant is approximately 1.0 MBtu/bbl. The cost of a 15 percent energy savings is then about $1.80/MBtu, and the cost of a 25 to 30 percent savings is $2.70/MBtu (assuming a 6 percent discount rate).
3. Petrochemical Plant (Nelson, 1989, personal communication, 1991). Use the data shown in Table 22.7 (Chapter 22). While no details of cost versus savings are available for specific projects, the data give a rough indication of the capital cost per unit of energy saved. For recent years an approximate figure is $1.25/MBtu assuming a 6 percent discount rate and a 5-year project life. Assume this investment cost applies to an overall energy savings in the neighborhood of 15 percent.
4. Overall Summary. Assume the results above apply generally to the industrial sector. Because actual discount rates used for industrial investments are substantially higher than the 6 percent value assumed here, many of the investments that are possible still have not been made (e.g., see Table 29.2 for costs at a 30 percent discount rate). To estimate net costs and CO2 reductions, use the current fuel-weighted price of industrial energy of approximately $3.00/MBtu and the current total of 1150 Mt CO2 from industrial fuel combustion (Edmonds and Ashton, 1989). For a 15 percent overall energy savings the average cost for the three cases above is $1.35/MBtu. The fuel-weighted average CO2 emissions are 0.07 t/MBtu. This yields a net cost of -$24/t at an overall CO2 reduction of 173 Mt/yr. If a real fuel price escalation were assumed, the net cost savings would be still greater.
For a 30 percent overall energy savings, extend the (nonlinear) steel plant data from 25 to 30 percent and average that cost with the refinery data (which shows a generally linear slope). The average investment cost is then about $3.05/MBtu. Use this as a rough estimate. The net cost is then 5 cents/MBtu, assuming no price escalation for the average fuel mix. This is equivalent to about $1/t CO2 at an overall reduction level of 345 Mt. With fuel price escalation the net cost would be slightly negative for a 6 percent discount rate, indicating an overall savings from the energy efficiency investments.
1. Throughout this report, tons (1) are metric; 1 Mt = megaton = 1 million tons.
Ayres, R. U. 1990. Energy conservation in the industrial sector. In Energy and the Environment in the 21st Century: Proceedings of a Conference at Massachusetts Institute of Technology. Cambridge, Mass.: MIT Press.
Barakat and Chamberlin, Inc. 1990. Efficient Electricity Use: Estimates of Maximum Energy Savings. Report EPRI CU-6746. Prepared by Barakat and Chamberlin, Inc. for Electric Power Research Institute. Palo Alto, Calif.: Electric Power Research Institute.
California Energy Commission. 1990. Staff Testimony on qualifying Facilities/Self-Generation Forecast. Docket No. 88-ER-8. Sacramento, Calif.: California Energy Commission.
Edmonds, J., and W. Ashton. 1989. A Preliminary Analysis of U.S. CO2 Emissions Reduction Potential from Energy Conservation and the Substitution of Natural Gas for Coal in the Period to 2010. Report DOE/NBB-0085. Washington, D.C.: Office of Energy Research, U.S. Department of Energy.
Electric Power Research Institute. 1989. EPRI 1989 Technical Assessment Guide, Electricity Supply. Report EPRI P-6587-L. Palo Alto, Calif.: Electric Power Research Institute.
Electric Power Research Institute. 1990. New push for energy efficiency. EPRI Journal 15(3):4–17.
Larsen, W. G. 1990. Energy Conservation in Petroleum Refining. Ph.D. dissertation. University of Michigan, Ann Arbor.
Larson, E. D., and R. H. Williams. 1985. A Primer on the Thermodynamics and Economics of Steam-Injected Gas Turbine Cogeneration. Report PU/CEES 192. Princeton, N.J.: Princeton University.
Nelson, K. C. 1989. Are there any energy savings left? Chemical Processing (January).
RCG/Hagler, Bailly, Inc. 1991. Industrial Cogeneration Markets. RCG/Hagler, Bailly, Inc., Washington, D.C. January 29. Memorandum to E. Rubin, Carnegie-Mellon University.
Ross, M. 1987. Industrial energy conservation and the steel industry of the United States. Energy 12(10/11):1135–1152.
Ross, M. 1990a. Modeling the energy intensity and carbon dioxide emissions in U.S. manufacturing. In Energy and the Environment in the 21st Century: Proceedings of a Conference at Massachusetts Institute of Technology. Cambridge, Mass.: MIT Press.
Ross, M. 1990b. Conservation supply curves for manufacturing. In Proceedings of the 25th Intersociety Energy Conversion Engineering Conference. New York: American Institute of Chemical Engineers.
U.S. Department of Energy. 1984. Industrial Cogeneration Potential (1980–2000) for Application of Four Commercially Available Prime Movers at the Plant Site. Report DOE/CS/40403-1. Washington, D.C.: U.S. Department of Energy.