Process oriented industrial classification based on energy intensity

Applied Thermal Engineering 26 (2006) 2079–2086 www.elsevier.com/locate/apthermeng

Process oriented industrial classification based on energy intensity Asfaw Beyene *, Annika Moman Department of Mechanical Engineering San Diego State University, San Diego, CA 92182, United States Received 21 September 2005; accepted 8 April 2006 Available online 13 June 2006

Abstract The Standard Industrial Classification (SIC) code has been used in the United States to categorize manufactured products. This classification is useful for many commercial and business related applications. However, when checked against energy use data collected through the Industrial Assessment Center (IAC) of San Diego State University (SDSU), the energy use profile of the SIC groups showed objectionable deviations within each group, rendering the code less than useful for overall energy based grouping and analyses. In this paper, a Process Oriented Energy Intensity Classification (POEIC) is introduced. When examined with data from about 270 plants of the IAC, this new classification offered more coherent and consistent energy use profile with smaller standard deviation for the selected energy intensity parameters than that of the SIC code. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: SIC; POEIC; Energy; Manufacturing; Process; Classification

1. Introduction The Standard Industrial Classification (SIC) code is used to categorize products based on the primary final output of a manufacturing plant [1,2]. This code does not discriminate plants by the manufacturing process or by the energy use profile such as thermal to electric ratio, energy intensity, electric load factor, etc. In fact, to date there is no such classification of manufacturing plants by their energy use to collectively address similarly energy intensive plants for heat recovery, efficiency, peak shedding, incentives, demand management, and other energy related purposes. Data from 370 energy assessment visits by the Industrial Assessment Center (IAC) at San SDSU were used to establish an energy use pattern among manufacturing plants in Southern California. The plants are defined as small and medium in size with energy bills ranging from $100,000 to $2.5 million per year, with less than 500 employees, and less than $150 million in annual sales. The data *

Corresponding author. E-mail address: [email protected] (A. Beyene).

1359-4311/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2006.04.019

collected at each of these assessments provide an excellent overview of manufacturing plants and their energy usage. All data are anonymous and plants are tracked by assessment number only. The IAC data were compiled and energy profiles were determined using energy bills and major plant equipment. All electric heat sources were converted to therms to accurately depict plant thermal needs. Of the 370 plants assessed, 270 contained the information needed for this analysis. The remaining incomplete sources with data were ignored or used only partly. The thermal to electric (T/E) ratio was determined for each manufacturing plant. This ratio provides insight to the process and energy requirement of the plant by comparing thermal and electric needs. Generally a T/E value of 5 or higher indicates a good candidate for a cogeneration system [3, p. 280]. This same value is also an indicator of waste heat recovery potentials. However, the T/E ratio does not differentiate between high and low grade heat and further investigation into the process is required to determine the quality of the waste heat. To categorize manufacturing energy use, 29 Process Oriented Energy Intensity Classifications (POEIC) were

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

Table 1 Major SIC codes and the recommended processes oriented classification SIC

Description

Process

20 22 23 24 25 26 27 28 29 30 32 33 34 35 36 37 38 39

Food and kindred products Textile mill products Apparel and other textile products Lumber and wood products Furniture and fixtures Paper and allied products Printing and publishing Chemicals and allied products Petroleum and coal products Rubber and misc. plastic products Stone, clay, and glass products Primary metal industries Fabricated metal products Industrial machinery and equipment Electronic and other electric equipment Transportation equipment Instruments and related products Miscellaneous manufacturing industries

Assembly Baking Boiling Bonding Chilling Curing Cutting Drying Extrusion Grinding Heat treating Incineration Injection molding Lamination Machining Melting Mixing Molding Painting Plating Press Press, punch Printing Sintering Soldering Welding

T/E Ratio and Energy Intensity by SIC 40.00

1800

35.00

1600 1400

30.00

1200

25.00

1000 20.00 800 15.00

600

10.00

400

5.00

200

0.00

0

Energy Intensity (kWh/$1000)

identified. The total energy use, electric and gas, was converted to kilowatt-hours, and divided by the total sales. This value assigns ‘‘intensity’’ to the process which is tied to both the technique and the product produced. Annual sales was collected and used to determine approximate energy intensities. Sales numbers from plants assessed prior to 2005 were corrected using the consumer price index to 2005 dollars. With the T/E ratio and energy intensity parameters determined for each site, the plants were grouped by major SIC and major processes. Determining the major process of plant involved reviewing the plant background. Once all reports were screened, some processes were combined to a reasonable number of categories allowing meaningful analysis. The result gave 27 major processes covering the eighteen major SIC codes. The SIC code has been recently replaced by the North American Industrial Classification Standard (NAICS) code. Changes of this reclassification are basically structural and in no way affected the analyses or results of this study. Table 1 identifies the major SIC and process categories used in the analysis. The T/E ratio and energy intensity for each major process and SIC categories were averaged. Figs. 1 and 2 depict the relationship between T/E ratios and energy intensity by SIC and by process, respectively. The process categories are arranged in descending order of energy intensity. Both figures show that the two values follow a similar pattern. One significant deviation exists in the process graph. Manufacturing of ice, ice cream, or other products by a process identified as ‘‘chilling’’, have high energy intensity but low

T/E Ratio

2080

20 22 23 24 25 26 27 28 29 30 32 33 34 35 36 37 38 39 SIC avg T/E Ratio

avg Energy Intensity

Fig. 1. Energy intensity and T/E ratio by major SIC.

T/E ratio. If the chilling could be accomplished using absorption chillers, or other similar technology, the T/E ratio of a ‘‘chilling’’ process would increase in line with the other processes of similar energy intensity. The process with the highest T/E ratio and energy intensity is incineration. The values are approximately one-and-a-half times that of the next process, drying. One process not visited by the IAC is plasma gasification; a vitrification process which uses a catalyst to gasify carbon based waste material in the absence of oxygen [4,5]. The operating temperature ranges from 2000 °C to 3000 °C. The synthesis gas, also known as Syngas leaves the gasifier at high temperature and typically consists of a mixture of H2, CO, CO2 and N2 [6,7]. This high temperature process has a T/E ratio of approximately 55. The relationship between SIC and energy profile is important because major policies, funding opportunities, utility activities, etc., depend on this relationship. Manufacturing plants are categorized within SIC codes of 20–39 and targeted for various activities by all US Departments. However, as can be seen from the list of processes identified within the plants considered in this study, not all plants have energy intensive processes. Plants with processes such as assembly, printing, and machining rarely have large energy saving opportunities like those identified at plants with processes like drying, curing, heat treating, and extrusion. This process based classification in fact allows making distinctions even within the same single manufacturing plant. For example, within SIC code of 30, the most frequently appearing category in our data, there are 14 different processes. The most common are injection molding, extrusion, molding, and pressing, but there are also plants whose major processes are assembly, soldering, welding, and painting. This disparity makes it difficult to identify with certainty which plants have energy intensive processes without visiting the plant, especially in the small and medium-sized plants. The top four most frequent SICs (30, 34, 20, and 33) and processes (mixing, extrusion, press, and machining) were examined more closely to determine consistency in T/E

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

2081

2500

35 30

2000 1500

25 20

1000

15 10

500

5 0

in ci n

er at i dr on y ch ing i m lling el p tin ex lati g n la trus g m io in n at cu ion gr rin in g d m ing i he mo xing at l d tre ing si atin nt g e ba ring kin i n pr g j m es o s pr w ldin es eld g s, in pu g b o n ch i pr ling so inti ld ng er in m cu t g ac tin h g pa inin as int g se ing bo mb nd l y in g

0

Energy Intensity(kWh/$1000)

T/E Ratio

T/E Ratio and Energy Intensity by Processes 45 40

Process avg T/E Ratio

avg Energy Intensity

Fig. 2. T/E ratio and energy intensity by major processes.

Table 2 Standard deviation by SIC and process SIC/Process

30** 34** 20** 33** Mixing Extrusion Press Machining

Number in sample

Standard deviation T/E ratio (%)

Energy intensity (%)

52 42 23 23 26 22 21 19

159 131 106 74 87 133 151 98

103 101 155 104 94 136 86 55

ratio and energy intensity within the categories. The T/E ratio and the energy intensity were plotted separately and regression analysis performed. The results are depicted in Table 2. The standard deviation percentage value was determined by dividing the standard deviation result from the regression analysis by the category average. While significant deviation exists in both SIC and process methods of grouping, the standard deviation by SIC is much higher; all but one of the eight values is over 100%. When the grouping was done by process, the standard deviation per-

centage was lower with five of the eight values below 100%. Auspiciously, inconsistencies of the process classification can be explained by variations within the details of the process, such as merging metal and plastic injection into a single process that gives high standard deviation. This can be corrected by reclassifying the process based on the injected material (Figs. 3 and 4). With such a low chance of predictability, SIC categories cannot be effectively used to screen and select manufacturing plants for targeted energy conservation measures. One example of the inconsistency in energy use within SIC category is two plants assessed within a single SIC of 2653, paper and allied products. One plant makes cardboard boxes and the other plant prints labels on cardboard boxes. The first plant has a T/E ratio of 19.15 and an energy intensity of 628 kW h/$1000 sales. The second plant has a T/E ratio of 0 and an energy intensity of 130 kW h/$1000 sales. Clearly the two plants, with the same SIC, are completely different from an energy perspective. By process, the first plant is ‘‘curing’’ and the second plant is ‘‘printing’’. These two categories give a much clearer indication of the energy usage of the two plants. Due to the large standard deviation within extrusion and press, these processes were examined more closely to

SIC 30 T/E Ratio Distribution 30.00 Std Dev = 159%

T/E Ratio

25.00 20.00 15.00 10.00 5.00 0.00

SIC 30 Fig. 3. T/E ratio distribution for plants with SIC 30.

2082

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

SIC 30 Energy Intensity Distribution 1800 1600

Std Dev = 103%

kWh/$1000

1400 1200 1000 800 600 400 200 0

SIC 30 Fig. 4. Energy intensity distribution for plants with SIC 30.

Extrusion T/E Ratio Distribution 30 Std Dev = 129%

T/E Ratio

25 20 15 10 5 0

Extrusion Fig. 5. T/E ratio distribution for extrusion with aluminum plants identified.

Extrusion Energy Intensity Distribution 3500

kWh/$1000

3000

Std Dev = 142%

2500 2000 1500 1000 500 0

Extrusion

Fig. 6. Energy intensity distribution for extrusion with aluminum plants identified.

see if a sub categorization was needed. Two types of extrusion were identified: plastic and aluminum. The circles on Figs. 5 and 6 indicate the plants that are aluminum extrusion. Separating these plants and recalculating the standard deviation resulted in a significant improvement in the results. Table 3 shows the standard deviation with extrusion and press separated into sub-processes. The standard deviation dropped between 3% and 83% points with the sub categorization. While the energy profiles within SIC and process categories varied significantly, a clearer trend emerged across the process categories, especially upon closer examination and

Table 3 Standard deviation by process (adjusted) SIC/Process

Mixing Extrusion (all) Extrusion (plastic) Extrusion (all) Press (all) Press (misc) Press (forging) Machining

Number in sample

Standard deviation T/E ratio (%)

Energy intensity (%)

26 22 14 8 22 19 3 19

87 133 115 65 151 110 68 98

94 136 63 120 89 86 32 55

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

categorization of the processes. The deviation within categories may be attributable to low sample numbers or missing information in the data collection. 2. Applications The process oriented classification is useful in identifying heat recovery, efficiency, peak shedding, incentives, demand management, and other opportunities including for policy formation and strategic planning. The benefits of energy efficiency improvements to a plant with a high T/E ratio are evident by the nearly linear increase in energy intensity; the greater the energy intensity the greater the economic benefit from any type of waste heat recovery. With the SIC system, identifying plants that fall into this category cannot be easily accomplished. Energy parameters such as T/E ratios and energy intensities are not consistent even for plants manufacturing the same final product. A previous effort to identify parameters such as energy usage per employee, electric rates, and energy conservation opportunities (types and savings) that may be consistent for any of the SIC codes offered no such uniformity in any of the cat-

2083

egories [2]. The plots showed no unique features that could be attributed to or generalized by SIC codes. Nine plants with processes in the top ten energy using categories were analyzed more closely for energy conservation opportunities. Although it ranks among the top ten energy usage process, chilling was not included because of its uniqueness mentioned earlier. The plants were selected randomly from within the categories with an attempt to provide a closer look of the process category for typical manufacturing plants. The nine processes were curing, drying, extrusion, grinding, incineration, lamination, melting, mixing, and plating. The results were then plotted to see if the waste heat recovery potential corresponded to T/E ratio and if the demand reduction potential related to the energy intensity of the process. Figs. 7 and 8 illustrate waste heat recovery and demand reduction potentials, respectively. The waste heat recovery potential generally followed the T/E ratio; in the case of the plant using a curing process, no waste heat recovery equipment was installed, whereas at the other high thermal plants, incineration and melting, some waste heat recovery was already taking place. This accounts for the large waste heat

500,000

40

450,000

35

400,000 350,000

30

300,000

25

250,000

20

200,000

15

150,000

10

100,000

5

50,000

0

WHR Potential (therms/yr)

T/E Ratio

T/E Ratio vs Waste Heat Recovery Potential 45

0 Curing

Drying

Extrusion

Grinding

Incineration Lamination

Melting

Mixing

Plating

Process T/E

Waste Heat Recovery Potential

Fig. 7. T/E ratio versus waste heat recovery potential by process.

70%

4000

60%

3500 50%

3000 2500

40%

2000

30%

1500

20%

1000 10%

500 0

0% Curing

Drying

Extrusion

Grinding

Incineration Lamination

Melting

Mixing

Process Energy Intensity

Percent Peak

Fig. 8. Energy intensity versus demand reduction potential by process.

Plating

Percent Peak Demand Reduction

Energy Intensity (kWh/$1000)

Energy Intensity vs Demand Reduction Potential 4500

2084

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

recovery potential at the curing plant. The demand reduction potential generally follows the energy intensity although this data is not complete as the energy assessments do not look for demand reduction opportunities specifically. Further investigation into the demand reduction opportunity needs to be conducted to verify this relationship. From the results of the above analysis, the potential benefits of process oriented classification can be observed. By determining the primary process in a plant, the potential for waste heat recovery and demand reduction can be identified. While the process classification will not provide accurate information about various potential opportunities, it is a significant step towards identification of such plants. The SIC classification system cannot provide as useful information. Two other applications for the POIEC are demand management through peak shedding and matching plants for energypark [8]. The first, demand management, is crucial to the health of the country’s energy distribution and generation systems. Demand management can be accomplished in several stages: permanent reductions, pre-determined short term reductions, and short notice response. Using the energy classification of manufacturing plants, utility and energy management personnel can target plants for each of the three stages. Energy intensity, electric load factor, and even waste heat recovery potential are all factors to be examined when determining peak shedding opportunities. Low electric load factors are indications of processes that surge at certain times; this can be a process like injection molding where the equipment cycles, or it can be a poorly designed system where all the equipment comes on at one time without regard for the demand. Waste heat recovery potential is usually associated with preheating combustion air, but it is also important for cogeneration applications which reduce demand on the electric grid. One specific energy conservation opportunity identified for six of the seven plants examined in detail is variable frequency drives (VFDs). VFDs ramp down the energy use when a motor is idling and can be applied in almost all variable load manufacturing applications. In the six plants, VFDs could reduce demand by 1–54%. While not all energy usage is predictable with a VFD, it is in a process like injection molding, or grinding and some mixing, where the timing of the process does not vary. VFDs in these applications could lead to significant and reliable demand savings. Another application of the suggested classification is specific to CHPs, particularly in selecting anchor tenant for multiple plants [8]. Such a system can be optimized for maximum efficiency as stated above for an energypark scenario. A process classification allows an initial screening of manufacturing plants that meet the energy usage needs of the park design. For example, when using the donor/ receiver model, the base plant of the park should be one with a high T/E ratio and energy intensity to provide a source of waste heat for the other partner plants. Attempt-

ing to select this plant by SIC code would be nearly impossible. The same applies for the receiver plants; the receivers should have T/E ratios that indicate thermal usage (and electric in some cases) meeting the output of the donor plant. Receiver plants with a summed T/E ratio larger than that of the donor plant may not be the best match for the energypark. T/E ratio and energy intensity are not sufficient indicators to select specific plants and further investigation should be conducted to determine the quality of the waste heat, especially the temperature and gas composition – pollution levels, or any other considerations. But, the number of plants investigated in detail is significantly reduced when selected by a process based energy classification system. The following example illustrates a specific application of POEIC. It has been shown that the concept of ecopark capitalizes on the benefits of Combined Heat and Power (CHP) by selecting plants with appropriate thermal to electric (TE) ratios and varying degrees of thermal load quality. Single manufacturing plants often have varying TE ratios during the manufacturing cycle. This requires bypassing and dumping excess energy, diminishing efficiency of the system. However, a CHP system that supports energypark – several TE matched plants, can operate at or near maximum efficiency at all times [8]. It is important to design a process by which partners of an energypark are selected. As stated above, neither SIC nor NAICS account for the manufacturing process or the energy profile of the plant including the TE ratio, energy intensity, and electric load factor. The two general options for plant selection are: to set up a grouping of similarly sized plants with different processes with multiple donors and receivers, or choose one base plant with a large amount of waste heat as a singular donor and several smaller plants as receivers. Resource streams form the second level of screening for potential associate plants. The example here demonstrates the second option. The base plant is a plasma gasification reactor (PGR). The PGR uses a municipal solid waste feed. The material feed can also be automobile shredder residue, industrial waste, biomass, or coal. The intense heat, 2500 °C, destroys chemical bonds in the material. Inorganic elements and compounds form a glassy slag that is discharged from the bottom of the reactor. Organic and gaseous compounds are discharged as a product called syngas. Due to the intense heat in this process, pollutant formation, primarily dioxins, is very low. Recovered Energy Inc. (REI) claims that the conversion of waste to salable product is 99%, REI web site [6]. The process may indeed qualify as a renewable energy source which would eliminate the regulatory hurdles normally associated with CHP systems [8]. For this reason, plasma gasification is an ideal candidate for an energypark. In fact, there is one existing full scale PGR plant in Utashinai, Japan. PGR has a very high TE ratio and energy intensity identified in the analysis of manufacturing plant. The first

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

grouping of parasite plants need to have TE ratios greater than one at a minimum with higher values, say over ten being more desirable, Thumann et al. [3]. The selection can be narrowed targeting the use of PGR byproduct such as slag, i.e., the second grouping of parasite plants targets use of the low grade heat available from cooling the slag material. A combination of TE ratios and waste/resource streams are used to identify ideal matches. The primary feed of the Utashinai plant is shredded automobile residue at a capacity of 165 tons per day. It operates two reactors generating enough syngas to power a 7900 kW steam turbine. The low pressure steam from the turbine is used for district heating. The system combusts the syngas in an afterburner at high temperature to break down dioxins. The hot gas is then put through a boiler where high pressure high temperature steam (40 atm, 400 °C) is generated. The gas must be cooled rapidly to prevent the formation of more dioxins which form between 450 and 250 °C; the faster the gas passes through this region the less dioxins will be formed, Bournemouth web site, [9]. After the boiler the gases pass through a cooling tower and then a bag filter which uses an alkali substance to capture any reformed chlorides and sulfides. After filtering the gas is exhausted to the atmosphere. Selecting the first group of parasite plants focuses on the waste heat available from the PGR. Considerable heat is available from the hot syngas leaving the PGR; the syngas must be cooled before it can be cleaned. Additionally, heat is available from the combustion of the syngas. Plants with high thermal requirements, high TE ratios, were selected. The processes considered were among the top ten energy intensive plants. In this case, the construction material industry was an obvious starting point within the processes due to the slag byproduct from the PGR. The three processes selected are drying, curing, and grinding. The selected plants produce asphalt and aggregates, concrete pipes, sand products, and brick. The data for these plants represent actual plants in the southern California, audited by the SDSU IAC. These plants also share resources and products amongst themselves; they can purchase raw materials in bulk such as mined dirt, thereby reducing costs. To utilize the heat available from water cooling the slag material, two agricultural plants are added. Livestock feeds and dairy both need low temperature and pressure steam and hot water, primarily for sanitation. The material flow in the energypark is an important selling point of the concept. The asphalt, brick, and concrete plants can all utilize parts of the glassy slag that is produced from the inorganic material in the PGR. Additionally, the aggregates produced by the asphalt plant can be used by the concrete plant and the dirt that is unusable can be transferred to the sand plant. The sand plant can sell some of its product to the concrete plant and can transfer the clay that must be separated from the sand to the brick plant. The syngas cleaning process removes sulfur from the gas which can be converted to fertilizer grade sulfur for the

2085

livestock feeds plant. Additionally, the biomass waste from both livestock feeds and the dairy is a suitable fuel for the PGR. The syngas leaves the PGR and passes through a heat exchanger to preheat the combustion air for the asphalt, brick, and sand plants. After the syngas cools to 132 °C1, the gas passes through a cleaning system. This cleaning system is a proprietary development of REI (REI web site) and uses off the shelf technology. The gas first passes through a cyclone where particulates are removed. These particulates, mostly metals with melting points below 1250 °C, are injected into the molten glass (slag). Once the gas is cleaned it is combusted and the exhaust gas turns a gas turbine. Two companies currently manufacture turbines that can run on low Btu synfuel, GE and Siemens. A heat recovery steam generator is used to produce steam for curing concrete. The slag leaves the bottom of the PGR and flows through a trough and is quenched. The trough is water cooled and the water is used by the livestock feed plant and dairy for sterilization and cleaning. The glassy portion of the slag forms glass fragments which allow easy separation of the metals. The salable products from the PGR are the fertilizer grade sulfur produced in the gas cleanup equipment, the metals that can be separated from the slag, and the glassy slag can be used as construction material. After selecting the energy system best suited for the energypark, the system performance must be analyzed. The total capital investment for the cogeneration system is $45,853,544. The total energy use of the plants is taken from field visits, and a standard industrial rate for southern California is used: $0.10/kW h, $10/kW, and $0.70/therm.2 The electric offset is calculated from the gas turbine power output and the gas offset is calculated from the heat recovery steam generator and the heat exchanger for the plant combustion air. The annual savings from energy costs not purchased from the grid amount to $7,893,288. The cost does not account for generating the necessary energy from the energypark system. The stream costs are used to determine the cost of the energy generated by the PGR system. The electric generation cost is $854,007 annually; this is significantly less than the grid electricity offset cost of $4,607,200. The cost of generating the thermal energy is $6,920,147, which is twice as high as the gas offset cost of $3,286,088. Additionally, the cost of generating the syngas is assumed to be $2/MMBtu and the cleaning is assumed to cost $1/MMBtu which is added to the energy cost of the PGR system. The total annual generated energy cost from the PGR system is $11,438,526. This results in an increased annual energy cost of $3,545,238. With the average cost of electricity and natural gas in southern California, generating electricity and thermal energy with a PGR system is not economically feasible. Tipping fees that would be avoided by the manufacturing plants for their 1 2

Temperature specified by Recovered Energy Inc. cleaning system. Based on the author’s experience with industrial rates.

2086

A. Beyene, A. Moman / Applied Thermal Engineering 26 (2006) 2079–2086

waste fed into the PGR are not included. Tipping fees in the US currently run at about $15/ton. A PGR with a 12 mton/h feed running year round would save $1,576,800. This would reduce the increased energy cost to $1,968,438 annually. In Japan, tipping fees run $100/ mton and the annual savings of the same size PGR are $13,140,000. This cost, coupled with electricity rates almost double those in the US, result in a net savings of $16,909,799. The only state in the US with similar tipping fees and electric rates is Hawaii; they are currently investigating a PGR for up to 1000 tons of waste per day Erkman [9]. With the tipping fees and energy costs in Japan, the payback of the PGR system is almost 7 years. This assumes a project cost of $45,853,544 and lifetime operating and maintenance costs of $82,536,379. Additional factors that could make the PGR more economical in the US are NOx credits and ‘‘green’’ incentives. Rebates for electricity generated, similar to those available for solar, wind, and fuel cell power, would significantly reduce the installed cost of the system. Also, avoided environmental remediation costs are not included. These represent the cost of cleaning up a landfill, the contamination of groundwater, or any other environmental costs associated with waste. An extended exergy accounting of the energypark system, including the costs of environmental impact, transportation, and other losses in all parts of the process has been presented [10].

gram like San Diego Gas and Electric’s (SDG&E) kWickview, would allow this type of analysis. Many plants have a real time metering capability and load factors that could be determined and used to establish these parameters. With the POEIC, manufacturing plants benefiting from comprehensive energy efficiency and demand management programs could be effectively targeted. With a process identified, the T/E ratio and energy intensity of the plant can be estimated with nearly 80% accuracy. With those values known, the waste heat recovery and demand reduction potentials can be quantified. Clear trends in energy usage could be seen and compared locally and nationally to determine areas for improvement. Unique situations requiring attention could also be isolated geographically. Development of an energypark or other cooperative scenario based on energy usage would be more easily accomplished. In summary, POIEC will provide an excellent tool for industry, local and national agencies, and utilities to improve energy efficiency and resource planning. Acknowledgements The authors would like to appreciate SDG&E for encouraging and financially supporting the research project that led to the publication of this paper. Guy Hansen and Donald Wiggins in particular deserve our special thanks.

3. Conclusion References Manufacturing energy use classification is useful to target plants for conservation opportunities, for policy developments to promote efficient energy use, or to promote newer and progressive concepts such as energyparks. A core group of primary processes should be identified with a sub-level added if further differentiation is necessary. This type of energy based classification will facilitate targeted energy efficiency programs and provide a clear starting point for any manufacturing sector energy projects. This in turn will help design optimum systems with superior efficiencies employing known and accepted design procedures. This new classification, POEIC could consist of three numbers, the first addressing the T/E ratio, the second addressing the energy intensity involved in the process, and the third a load factor calculation. Some manufacturers may be hesitant to provide detailed information about their process to allow specific calculations, but with a detailed study of each type of process prior to implementation, categorizing plants with a 0–9 value in the three areas would be achievable. Unfortunately, a load factor comparison was not possible with the IAC data presented above. Daily energy profiles, such as those available with a pro-

[1] Standard Industrial Classification (SIC) Index. Available from: , Numerical Headings Index; accessed March 21, 2005. [2] A. Beyene, Energy Efficiency and Industrial Classification, Energy Engineering 102 (2) (2005) 59–80. [3] A. Thumann, P. Mehta, Handbook of Energy Engineering, fifth ed., Fairmont Press, Lilburn, GA, 2001. [4] J. Balgaranova, Plasma chemical gasification of sewage sludge, Waste Management & Research 2 (2003) 38–41. [5] C. Butcher, Siemens takes the Lead in Gasification in Europe, Turbomachinery International, May/June 2004 [online magazine]. Available from: ; accessed March 21, 2005. [6] Recovered Energy Inc., Plasma Gasification Overview. Available from: ; accessed November 1, 2004. [7] Westinghouse Plasma Corporation. Plasma Gasification. Available from: ; accessed October 12, 2004. [8] A. Beyene, Combined heat and power as a feature of energypark, ASCE Journal of Energy Engineering 131 (3) (2005) 173. [9] S. Erkman, Industrial ecology: an historical view, Journal of Cleaner Production 5 (1997) 1–10. [10] Moman A., Beyene A., Exergy analysis of energypark, unpublished paper (under review).