Many manufacturers have the ability to substitute the consumption of one fuel for that of another when the economic conditions call for making such a change. The ability to switch can be limited not only by the technical considerations of onsite boilers and heaters, but also the practical considerations of

• Federal, state, and local environmental restrictions;
• Available supply and transportation of the fuel;
• Contracts in which the manufacturer is committed to purchasing a certain amount of the fuel, regardless of what has transpired since the contract was agreed upon.

The Manufacturing Energy Consumption Survey (MECS) measures fuel-switching capacity by asking manufacturers to account for both technical and practical considerations simultaneously. Manufacturers were not to consider the relative prices of the fuels when they reported their switching capability. They were instructed to limit their ability to what can be done within a time-period of 30 days and without major modifications to their existing equipment.

A fuel may be completely or partly nonswitchable. In the case of it being completely nonswitchable, it often means that the configuration of the plant’s equipment would not allow another fuel to be used in place of the one currently being consumed. Even when an establishment has the ability to switch a portion of its consumption from one fuel to another, it could have a certain amount for which it cannot substitute. The term used to describe both types of situations is, “nonswitchable minimum requirement.” For example, a plant may use natural gas in both a boiler and specialty oven used to dry paint. In the boiler, all fuels used may be completely interchangeable. However, in the oven, using something other than natural gas may alter the pigment of the paint. In that case, the natural gas would be nonswitchable. Other reasons a fuel may be partly nonswitchable are the practical considerations such as in-place take-or-pay contracts with a supplier or environmental restrictions that limit the amount the establishment can consume during a specified time.

The MECS measured fuel-switching capability of a selection of the most common purchased energy sources (Figure 1). The total fuel consumption accounted for by those energy sources was 10,512 trillion Btu, approximately 65 percent of all fuel use in manufacturing. Most of the fuels for which the MECS did not measure fuel-switching were byproducts, waste products, and otherwise rarer substances. The interest in measuring an establishment’s ability to switch out of those substances is minimal as many of those fuels are produced on-site or acquired through means other than open-market purchases.

Because of the relatively higher amount of not-ascertained switching in 2002 compared to 1994 (18 percent to 6 percent), comparisons are problematic. However, the ratio of total switchable fuel consumption to the sum of all ascertained switchable and nonswitchable fuel of all measured fuels allows a useful comparison without the not-ascertained amount. In 1994, that ratio was 25 percent while in 2002, the ratio apparently dropped to 20 percent. While not entirely conclusive, this result indicates that there may have been some decline in fuel-switching capability between 1994 and 2002.¹

Table 1. Fuel Switching Percentages Based on Total Consumption and Sum of Switchable and Nonswitchable Amounts, 1994 and 2002
	Year	Total Consumed	Switchable	Not Switchable	Not Ascertained	Switchable	Not Switchable
		(Trillion Btu)	Percentages Based on Total Consumption)			Percentages Based on Sum of Switchable and Nonswitchable Amounts)
Electricity Received	2002	2,918	2.7	78.2	19.1	3.4	96.6
	1994	2,742	1.9	93.4	4.7	2.0	98.0
Natural Gas	2002	5,805	18.8	62.7	18.5	23.1	76.9
	1994	6,135	27.8	64.8	7.4	30.0	70.0
Distillate Oil	2002	142	20.1	55.7	24.2	26.5	73.5
	1994	152	18.0	68.2	13.8	20.9	79.1
Residual Oil	2002	208	41.4	44.6	14.0	48.1	51.9
	1994	441	45.2	51.3	3.4	46.9	53.1
Coal	2002	1,356	30.2	58.0	11.8	34.2	65.8
	1994	1,218	46.1	52.3	1.6	46.8	53.2
LPG	2002	96	22.8	32.6	44.5	41.2	58.8
	1994	94	41.8	52.1	6.1	44.5	55.5
Total	2002	10,525	16.3	65.7	18.0	19.9	80.1
	1994	10,781	24.0	70.1	6.0	25.5	74.5
Note: Values in the table are compiled from fuel switching tables 10.2-10.12 (http://www.eia.doe.gov/emeu/mecs/mecs2002/data02/shelltables.html) and converted to Btu.    Sources: Energy Information Administration, Manufacturing Energy Consumption Survey -- Tables 10.2, 1994 and 2002; 10.4, 1994 and 2002; 10.6, 1994 and 2002; 10.8, 1994 and 2002; 10.10, 1994 and 2002; 10.12, 1994 and 2002.

Referring to both Figure 1 and Table 1, natural gas, as the fuel with the most total consumption, accounts for most of the switchable consumption on a Btu basis in 2002. However, all but one of the energy sources measured (electricity) has higher rates of switchability on a percentage basis. Electricity and natural gas have the highest non-switchable percentages and also account for the greatest share of the nonswitchable total among all fuels.

Nonswitchable Minimum Requirements, Maximum Possible, and Discretionary Fuel Use

The MECS measured the nonswitchable minimum requirements of a fuel and the maximum possible if all possible switching into that fuel took place. By comparing those limits with actual consumption, one can compute a measure of discretionary fuel use. The discretionary-use rate is a measure, in percent, of the extent to which manufacturers elected to consume discretionary quantities of a given energy source. That is, manufacturers will consume certain fuels above the absolute minimum nonswitchable consumption. A measure of discretionary fuel use can be defined as:

where

USE is the discretionary-use rate of a given energy source;

ACT is the actual consumption of that energy source;

MIN is the minimum consumption, which would have been achieved if all ascertained switching from that type of energy had occurred, and

MAX is the maximum consumption, which would have been achieved if all ascertained switching into that type of energy had occurred.

The higher the ratio USE is, the closer the actual consumption was to the maximum possible consumption of that fuel, given all possible fuel-switching into the fuel and the same level of operations. Thus discretionary fuel consumption is a measure of preference for one fuel over another.

Table 2. Discretionary Fuel Use Rates and Ratios of Minimum to Maximum Consumption, by fuel, 2002 and 1994
	Discretionary Fuel Use Rate (Percent)		Minimum/Maximum (Percent)
Energy Source	1994	2002	1994	2002
Electricity Receipts	25.2	37.1	92.8	93.0
Natural Gas	73.4	69.5	65.6	75.0
Distillate Fuel Oil	3.1	5.0	12.2	16.5
Residual Fuel Oil	19.8	20.9	19.3	22.9
Coal	80.3	87.0	48.5	67.7
LPG	5.3	5.0	6.9	14.4
Note:Values in the table are compiled from fuel switching tables 10.2-10.12 (http://www.eia.doe.gov/emeu/mecs/mecs2002/data02/shelltables.html) and converted to Btu.    Sources: EIA, Manufacturing Energy Consumption Survey -- Tables 10.2, 1994 and 2002; 10.4, 1994 and 2002; 10.6, 1994 and 2002; 10.8, 1994 and 2002; 10.10, 1994 and 2002; 10.12, 1994 and 2002.

As can be seen in Figure 2 and Table 2, the discretionary fuel rate for coal and natural gas are quite high, indicating strong preferences for those fuels. Conversely, Table 2 shows a strong reluctance to use distillate fuel oil and LPG. The pattern of discretionary fuel use rates among the fuels in 2002 is markedly similar to the pattern found in 1994. However, it does appear that preferences for electricity receipts and coal have increased since 1994. In the case of electricity receipts, the increase in the discretionary fuel use rate mirrors the decrease in ratios of electricity generated onsite to electricity receipts (15.7 percent in 2002, down from 17.7 percent in 1994)². The increase in preference for coal may be a direct result of the rise in prices of other energy sources.

Although the discretionary fuel rates have not changed substantially in most cases, it appears that the discretionary ranges themselves have decreased (Table 1). This finding is verified by the fact that the ratios of the switchable amounts to the nonswitchable have decreased for the total and all the fuels, except distillate and residual fuel oil (See data table below Figure 1).

Fuel-Switching Flexibility

Another measure that is useful in evaluating fuel-switching capacity is flexibility. Flexibility in fuel-switching is a measure of redundancy in the alternatives available to a manufacturer to switch out of a fuel. Flexibility will always be measured as percentage greater than 100 as it is defined as the sum of the quantities of alternative fuels, divided by the total amount of switchable fuel. For example, an establishment reports that it consumes 15,000 thousand cubic feet (MCF) of natural gas, and that the equivalent of 10,000 MCF of which could have instead been consumed as other fuels. The establishment also reports that 7,500 MCF could have been switched into residual fuel oil, and 5,000 MCF could have been switched into LPG. If no other alternatives were reported, than the establishment’s flexibility in switching to natural gas was 125 percent (computed as (7,500 + 5,000) ÷10,000). The establishment essentially has flexibility in choosing how it would use residual fuel oil and LPG to substitute for natural gas over the year if it chose to do so.

Table 3 shows the measures of flexibility of energy sources over all manufacturing in 2002 and 1994.

Table 3. Fuel Switching Flexibility by Fuel Type, 1994 and 2002
Energy Source	Flexibility (Percent) by Year
	1994	2002
Natural Gas	126.27	116.37
Electricity Receipts	145.17	136.48
Coal	152.13	127.81
Residual Fuel Oil	152.90	129.23
Distillate Fuel Oil	138.74	143.33
LPG	N/A3	N/A3
Note: Values in the table are compiled from fuel switching tables 10.2-10.12 (http://www.eia.doe.gov/emeu/mecs/mecs2002/data02/shelltables.html) and converted to Btu.     Sources: Energy Information Administration, Manufacturing Energy Consumption Survey -- Tables 10.2, 1994 and 2002; 10.4, 1994 and 2002; 10.6, 1994 and 2002; 10.8, 1994 and 2002; 10.10, 1994 and 2002; 10.12, 1994 and 2002.

As can be seen in Table 3, flexibility has decreased for all fuels since 1994, except distillate fuel oil. Put together with the decrease in the overall ability to switch fuels and the smaller discretionary ranges, it does appear that manufactures have reduced their capacity to switch fuels from 1994 to 2002.

Possible reasons for the observed decrease in switchability since 1994 are:

• The use of long-term contracts in natural gas purchasing;
• Continuing concerns about complying with environmental laws, both in storage of fuel oil and LPG and in emissions, have continued to favor the use of natural gas over other combustible fuels;
• An overall decline in the use of fuel oils for fuel, while relatively unswitchable electricity receipts account for a greater share;
• The relative inexpensiveness of natural gas for most of the period between 1994 and 2002, coupled with its more desirable environmental characteristics.

As energy prices have risen dramatically since 2002, it will be interesting to see whether manufacturers have increased their ability to switch fuels in response.

Other Observations

• Examination of fuel-switching capability of natural gas in 2002 by value of shipments categories and employment size show that the ability to switch increases on average from smaller to larger establishments until the final category when there is a sharp drop off. Thus, fuel switching capability out of natural gas seems to correlate with size, regardless of industry.
• The relative inability of chemical companies to switch⁴ may be due indirectly to their heavy use of natural gas as a feedstock. It might not be worthwhile for them to invest in capacity to switch out of gas since their operations are so dependent on it. Further, those manufacturers usually get lower-than-average prices due to their large usage.
• The available alternatives to the switchable fuels vary with the fuel in question. For example, there is a fairly wide distribution of alternatives available for natural gas: 38 percent can be switched to distillate fuel oil, 34 percent to LPG, and 22 percent to residual fuel oil. Yet, for residual fuel oil, 81 percent was switchable to natural gas, distillate fuel oil could replace 24 percent, and no other alternative could replace more than 10 percent of the total.

Endnotes

¹ The assumption used here is that the not-ascertained amount would breakout in the same proportions as the fuel consumption that was actually measured as switchable or nonswitchable. As the not-ascertained amount increases, so does the potential for bias and the uncertainty of that assumption.
² MECS Tables, Electricity: Components of Net Demand, 1994 and 2002.
³ Some of the necessary data was unpublished because of large sampling error.
⁴ MECS Table 10.2 Capability to Switch Natural Gas to Alternative Energy Sources, 2002.

Contact

Robert Adler
robert.adler@eia.doe.gov
Survey Manager
Phone: 202-586-1134
Fax: 202-586-0018

William Gifford
william.gifford@eia.doe.gov
Mathematical Statistician
Phone: 202-586-5931
Fax: 202-586-0018