Attendees:

OIAF: Mary Hutzler
Mark Rodekohr
Barry Cohen
Crawford Honeycutt
Andy Kydes
Alan Beamon
Jim Diemer
Cedric Britt

EEUISD: Transportation & Industrial Branch Lynda Carlson
Stephanie Battle
John Preston
Bob Adler
Mark Schipper
Cindy Magee
Mike Margreta

Due to the Energy Policy Act of 1992 (EPACT), the Manufacturing Energy Consumption Survey (MECS) has become a biennial collection, instead of triennial, starting with the 1994 collection. In addition, it is expected that the sample size will be expanded to accommodate data reporting on the basis of census division. Also provided will be counts of establishment types.

Each attendee from OIAF was asked to give a brief reason for participation at the meeting. In general, OIAF is interested in obtaining accurate manufacturing energy data that serve as input to its industrial models. Also, OIAF is seeking consistency in MECS estimates with those from other government surveys. Other topics of concern are the survey collections for cogeneration, demand-side management (DSM), and fuel switching between natural gas and residual fuel oil.

OIAF raised an important question about the large and expanding gap between MECS estimates of natural gas use in manufacturing and State Energy Data System (SEDS) estimates of industrial gas use. Between 1988 and 1991, the SEDS estimate of industrial gas use increased by almost one quad (quadrillion Btu), while the MECS estimate for manufacturing increased by less than ¼ quad.

This discrepancy may result in the way certain nonutility generators are classified. Nancy Kirkendall at the Office of Statistical Standards (OSS) is checking into getting consistent definitions. Possibly, some nonutility generators are coming on-line with gas, but such generators are not classified as part of the manufacturing establishment. Hence, the MECS would not measure this energy at the consumption end. By means of Form EIA-846D, the MECS makes an attempt to measure electricity generation from nonutility generators that are wholly owned by manufacturers but physically located offsite from the establishment. There is no comparable vehicle for natural gas, although manufacturers should be aware of how/how much gas is being used, even if they are not always aware of its source. At this point, the extent of the natural gas inconsistency is not known.

• Note: In a cc:mail document dated 5/9/94, Mr. French provided excruciating detail of the possible reasons for this discrepancy in natural gas reporting.

Mr. Honeycutt has developed modeling formulas for steam. However, the MECS does not try to pinpoint how much steam is generated because it is surmised that respondents would have trouble figuring the total amount of steam. Steam would be easy for respondents to double-count because it cascades throughout the plant. Also, the MECS does not allocate steam to end uses.

OIAF suggested doing a pilot study on steam with some plants, in concert with engineering editing tools to determine consumption.

Also, OIAF would like to see collected such "boiler plate" information as boiler sizes, types, and ages. Both Mr. French and Mr. Preston suggested using focus groups to obtain this information, rather than a survey.

Mr. Honeycutt stated that data suppression, to avoid disclosure for individual establishments, in the category of food and kindred products (SIC 20) was not helpful to his modeling efforts. An expected increase in MECS sample size of 7,000 to 10,000 cases should alleviate this problem.

OIAF is also interested in collecting the demand for both interruptible gas and firm gas. A volume cutoff could be used to target certain survey questions to certain levels of gas usage. As such, it would be necessary to coordinate definitions with other survey vehicles, such as Form EIA-176. Mr. Diemer offered his assistance in the coordination effort.

On every survey, the MECS team revamps the sample design. For the 1994 MECS, this design will include data reporting at the level of census division, which is compatible with OIAF's modeling. Another enhancement is the tabulation of establishment counts for major groups (two-digit SICs) at the national level. As requested in a meeting last March, Mr. Cohen and Mr. Honeycutt have provided a memo with 26 manufacturing industry groupings, based on SIC codes, that they would like to have estimated in the next MECS. The MECS will try to accommodate these groupings:

1. SIC 20
2. SIC 21
3. SIC 22
4. SIC 23
5. SIC 24
6. SIC 25
7. SIC 26
8. SIC 27
9. SICs 281, 282, 286, & 287
10. SICs 283, 284, 285, & 289
11. SIC 2911
12. SIC 29 except 2911
13. SIC 30
14. SIC 31
15. SICs 321, 322, & 323
16. SIC 324
17. SICs 325, 326, 327, 328, & 329
18. SIC 331
19. SIC 3334
20. SIC 33 except 331 & 3334
21. SIC 34
22. SIC 35
23. SIC 36
24. SIC 37
25. SIC 38
26. SIC 39

OIAF does not use all of the fuel-switching categories supplied in the 1991 MECS. It really only needs switchable amounts of energy between natural gas and residual fuel oil, plus the reason(s) that caused the switch. If price was, or could be, the reason for the switch, OIAF would like to know how large of a price differential must occur to cause switching.

In addition, OIAF feels that gas demand-side management will also be an issue of concern. It would like to see a "yes-no" question added to the survey if the establishment is doing anything for gas DSM.

OIAF would like to know what motors are hooked up to, and what their emission levels are.

OIAF would like more information collection on nonpurchased energy sources.

A second meeting was requested to further discuss the "requirements and priorities" list that OIAF handed out. This meeting was scheduled for Wednesday, May 11, 1994.

Attendees:

OIAF: Mary Hutzler
Barry Cohen
Crawford Honeycutt
Jim Diemer
Cedric Britt

EEUISD: Transportation & Industrial Branch Lynda Carlson
Stephanie Battle
Bob Adler
Mark Schipper
Cindy Magee
Mike Margreta

This meeting was a continuation of the MECS User Needs Meeting held May 4, 1994, at which time a "requirements and priorities" list was handed out.

The first topic of discussion was Mr. French's cc:mail document of 5/9/94 that provided reasons for discrepancies between MECS estimates of natural gas use in manufacturing and those of SEDS.

Another explanation for this difference is that there was a one-time adjustment made in the MECS manufacturing sector for the year 1988. This adjustment was attributable to the relationship between the Annual Survey of Manufactures (ASM), which feeds the MECS, and the Census of Manufactures (CM). The ASM has limitations because it is updated two years after the last CM is conducted. The 1988 MECS was done the year before the ASM was to be updated, thereby causing MECS data collection to occur at the least opportune time. Hence, adjustments were made to cover the CM non-mail file, which includes approximately 125,000 establishments (2% of value of shipments and receipts) which are in the manufacturing sector but not in the group that is mailed the CM. Therefore, these one-time adjustments were made in the 1988 MECS to account for this coverage shortfall. However, no similar adjustments were made to the 1991 MECS because it is now believed that there is fairly consistent coverage of all establishments. Consequently, if the 1988 MECS adjustments were too large, then the expected increase in the 1991 MECS natural gas data would appear too small and would not be comparable to the SEDS increase.

In another matter, Survey Form EIA-176 has now been revised so that it breaks out commercial consumption and industrial consumption, excluding nonutility power producers and consumers. In addition, electric power data on Form EIA-423 will now be collected and tabulated separately for utility and nonutility power producers. These updates should help explain the differences between the MECS and SEDS.

Another data discrepancy occurred in the area of onsite generation of electricity, as reported by the MECS and Form EIA-867. The two areas of largest differences are in chemicals and allied products (SIC 28) and petroleum refineries (SIC 29). For SIC 28 in 1991, the MECS shows 41 billion kilowatthours of onsite generation, while the 867 totals 67 billion kilowatthours. Differences in establishment classifications could account for the discrepancy, for example, the way in which the Bureau of the Census classifies a petrochemical plant. However, these types of data discrepancies should be discussed at another meeting.

Residual fuel oil and natural gas are OIAF's main concern regarding fuels involved in switching. One issue is figuring out how much of the potential was switched over the year and why the switching was occurring. Given summer-winter seasonality, was the switch caused by fuel availability or by price? It is critical for OIAF to understand why the switch happened. If the respondent answers price, then OIAF would like to know the price difference, possibly as expressed in numeric categories by ascending percentage of difference.

Mr. French graciously provided the following diagram as an aid in digesting the fuel-switching scenario:

In this diagram, the large oval represents the total fuel consumed by a manufacturer, and it encompasses one large circle representing consumption of residual fuel oil and another for natural gas consumption. Within each large circle, a slightly smaller circle (call it a medium-sized circle) depicts the switchable amount of the fuel. Within the switchable amount, most likely not all of the fuel will be switched. The small, shaded circle represents the amount of the fuel (residual fuel oil in the left-most circle, natural gas in the right-most circle) that was consumed in operations where a switch occurred during the year. Therefore, the two shaded circles add up to the total residual and natural gas used in operations when the manufacturer switched back and forth during the year. The two medium-sized circles added together represent the total of residual fuel oil and natural gas that is available to be switched back and forth, regardless of whether or not the manufacturer actually switched.

OIAF would like to know the ratio of the shaded circle to the medium-sized circle for each of the two fuels discussed. Also, how many times was fuel switched, and what percent of the time was fuel availability/price differential the deciding factor for fuel switching? During the time of the price differential, on average what was the price difference?

The MECS team suggested that some of this information could be gathered by means of a focus group or else a follow-up subsample to selected manufacturers.

See Addendum: Fuel Switching at the end of this memo.

OIAF asked about collecting information related to EPACT standards for motors. Mr. French replied that it was unlikely that the manufacturers, if asked, could supply this level of detail.

OIAF also asked about collecting emissions data from the manufacturing process. Mr. Schipper stated that he had a file from EPA on local emissions and that EPA has continuous monitoring, with data collection by U.S. states. Mr. French stated that the MECS will not become involved with emissions collection.

In addition, OIAF expressed interest in boiler types, specifically the size and age of the boiler. OIAF will prepare a list of questions on boilers, including their size categories and boiler retrofitting.

Also of interest is energy consumption during various stages of production. Mr. French stated that the MECS will not become involved with energy consumption in the various process stages of production.

In conclusion, OIAF will get back to the MECS team with some model questions on the topics discussed above.

On May 16, 1994, the MECS team met separately with Mr. Cohen to specifically explore OIAF's needs regarding fuel switching.

OIAF would like some combuster-specific information, such as what quantity was switched, or at least what proportion of capacity was switched. The MECS team is not certain that such information can reasonably be supplied by establishments. Possibly, the MECS team will consider using focus groups to determine if this questioning is practical.

OIAF would like to know:

1. the number of fuel switches during the year
2. the reasons for the fuel switch (via checklist)
3. the magnitude of the minimal price difference at which fuel switching would occur (via percentage categories).

The MECS will try to be accommodating.

Contact:

robert.adler@eia.doe.gov

Robert Adler

Project Manager

Phone: (202) 586-1134

Fax: (202) 586-0018