Setting Initial Targets for Performance Improvement

Performance improvement targets are important for an energy and environmental management program because they are clear statements of what management wishes to achieve and when. Targets mustmobilize people toward the achievement of management objectives, hence they must be:

• capable of being monitored;

• capable of maintaining standards.

The starting point for target setting is the knowledge of current position. Looking back to our tuna-canning case study, even with only monthly energy/production data available, we are still able to identify instances of good performance which may be used for establishing targets at the company level. However, we have said that targets must be measurable and capable of being monitored. For that purpose, we must use some of the tools provided by regression analysis in order to quantify performance results and compare them with given targets. At the outset of the discussion on target setting, we have to emphasize that there will be always at least two levels of targets:

(i) targets at the company level;

(ii) targets on the ECC's level.

The ECC's targets need to be broken down further and established for different products and for various energy types used at the ECC.

4.6.1 Setting Targets at the Company's Level

The targets at the company's level are usually aggregated and based on available monthly data on achieved production and raw materials and energy used over a period of at least least 12 months. Such data, when checked for adequacy, represent average performance over the selected period, and provide the basis for setting overall targets for performance improvement.

The realistic specific initial target can always be based on the best performance already achieved in the past. In this case, this can be a target line that passes through the three points representing best performance in the respected period, as shown in Figure 4.9. In Part I Chapter 3, we have already calculated regression equation for the best-fit line based on these three data points:

Again, before deciding firmly that this equation can present realistic (achievable) target, we must satisfy ourselves that particular performances at these three months are achieved under ordinary circumstances.

The 'best-fit' line shows the average performance over the corresponding period, and the target line provides a clear objective for performance improvement at any ordinary production output. It is important to note that the target cannot be just one figure, for instance one ton of steam per ton of product, because energy consumption cannot be fixed irrespective of actual production level. Therefore, we speak about the 'target line' which is expressed by target equation, so that for any production output that may occur, a corresponding energy target value can be calculated and compared with actual energy used.

As said earlier, environmental impacts are consequent to the energy, water and raw material use, therefore there is no need to set specific target equations for environmental issues in operations, because when energy, water and raw material use are at their lowest for the desired level of production, the resulting environmental impacts will also be at their lowest. Only for end-of-pipe treatment of effluents and discharges, can specific targets be set, that will be achieved by applying particular environmental abatement technologies.

2500

<3 1000

E ra ffi

55 500 0

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Production [t/m]

Figure 4.9 Setting Target Based on Best Past Performance

4.6.2 Setting Targets on ECC Level

The targets for ECCs are set in the same way as for the company as a whole. But when trying to establish targets for an ECC, we almost invariably stumble on the problem of a lack of data. We have set initial targets for the factory based on the aggregated data on energy and production, but reliable data on production and energy use at the ECC level are usually not available prior to implementing an EEMS project.

Later, when the energy and environmental management system provides data for an individual ECC, we will need to set individual targets for ECCs in the similar manner as the aggregated ones. The overall target and individual targets should not necessarily have the same value.

Why? It is because targets must always be established from the position of knowledge of the actual performance at respective ECCs. For a factory, this knowledge is derived from the given monthly data set and subsequent analyses. If similar data for an individual ECC are not available and if we set a target of say 10 % improvement across the factory, that may be easy to reach for some ECCs because their actual potential may be larger, but for others it may be impossible because their actual performance may already be quite good.

Usually, individual ECC targets can be established when the initial operation of the EEMS provides daily energy/production data for at least two to three months. These initial targets should be checked periodically for adequacy and reset if necessary.

4.6.3 Targets for Different Products

When a company or ECC produces more than one product, targets must be established for each product separately. At the first glance, it may look like a complex and complicated procedure, but actually it is quite simple. The principle for target setting is always the same as described above and it must always be based on measured data. An operational EEMS will provide daily data on energy and production and, if necessary, a further break-down into data by shifts. This enables the separation of energy and production data on a product by product basis. Data for individual products is consolidated into one scatter diagram, so we will have the same number of scatter diagrams as there are individual products where performance needs to be monitored separately. These scatter diagrams then provide the basis for performance assessment and target setting for individual products in the same way as has been already described.

ft □

-'nr"""'

□ /

Targe

Software 1: Target Setting for Performance Improvement

Figure 4.10 Flow Diagram of the Software for Performance Improvements Target Setting

To help illustrate target setting from past performance data, a sample software tool (see Part III Software 1) has been developed. It provides the tool for presenting the energy performance of the industrial process in a scatter diagram. Once entered, data can be analyzed and targets determined either as uniform linear reduction, or based on selected best past performance points, or based on new discretionary target points. As the final result, the software provides target line equation. The software flow diagram is shown in Figure 4.10.

4.7 Monitoring Energy and Environmental Performance

A monitoring procedure will need to be introduced to follow up and check regularly on the progress of performance improvements towards targets. The information on current performance level must be provided by the performance measurement system through continuous measurement of key performance indicators.

The usual monitoring practice in factories is to record measured data in a tabular form into log sheets. However, this is the least informative way of data presentation for the purposes of performance monitoring. A much more informative way of presenting data is by means of the already described scatter diagrams. This is particularly true for monitoring energy consumption and environmental impacts from manufacturing operations, where production output figures are set out as an independent variable.

There are also other means of data presentation, which may be more appropriate for performance monitoring where there is a direct input/output relationship of the variables involved. These include load profile diagrams, which present variations of load over a period of time (an hour, a day or a week) and performance control charts that emphasize variations around a given target value.

Any presentation style is acceptable as long as it flags substandard performance promptly and indicates the potential causes of performance deviations.

Software 2, provided within the Toolbox, entitled 'Performance Monitoring and Control', offers an example for visualizing data and structuring reports on performance (Fig. 4.11). The inputs are a target

Environmnetal Managemnt Targets
Figure 4.11 Flow Diagram of 'Performance Monitoring and Control' Software

equation determined earlier and data on performance measurement received from performance measurement system. At any moment of time a user can select the style of visualizing performance data, as an input to the performance evaluation of the process and an interpretation of the data pattern that should identify possible corrective actions. The software also provides a structured means for reporting performance results and providing proposals for corrective measures, as well as generating daily, weekly or monthly reports.

4.8 Verifying Performance Improvements - CUSUM Technique

The nature of energy performance improvement projects and their results (energy and other cost savings) is somewhat hypothetical because actual savings and performance improvements are defined as post-improvement consumption or costs, subtracted from agreed base line consumption, that would have occurred had the project not been implemented, where all other factors remained constant.

When this is applied against the changing operating environment, where other factors are not constant, it is not surprising that in many cases people actually wonder how to determine how much, if any, they have improved their performance. In fact, this is the question that has puzzled a number of managers we have worked with. Usually, they produce two annual data sets on production and energy and all kind of colorful time-series graphs, but the question: 'How much did we save on the year to year basis?' remains unanswered!

Let us have a look at a real set of the data (Table 4.3, Table 4.4) representing monthly energy use and production output over two consecutive years and a set of diagrams (Fig. 4.12) that the factory has produced in order to try and answer the question: 'How much did we save?'

Table 4.3 Year-On-Year Data for ECC 1

Month

Production

Electricity

Month

Production

Electricity

Units

MWh

Units

MWh

Apr-OO

8584

1003

Apr-Ol

8787

848

May-OO

10 877

1199

May-Ol

10878

1153

Jun-OO

10 880

1203

Jun-Ol

10918

1134

Jul-OO

10211

1175

Jul-Ol

9710

1059

Aug-OO

10 567

1195

Aug-Ol

11434

1227

Sep-OO

10 741

1155

Sep-Ol

10114

1124

Oct-OO

10 161

1108

Oct-Ol

10344

1126

Nov-OO

10 784

1205

Nov-Ol

10538

1135

Dec-OO

8728

1021

Dec-Ol

8228

1033

Jan-Ol

10 020

1068

Jan-O2

9778

1042

Feb-Ol

10 007

1071

Feb-O2

8146

962

Mar-Ol

implementation

Mar-O2

Table 4.4 Year-On-Year Data for ECC 2

Month

Production

HFO

Predicted

Difference

CUSUM

Units

l

m3

m3

m3

Apr-00

8584

172

Apr-01

8787

181

May-00

10 877

206

May-01

10 878

202

Jun-00

10 880

218

Jun-01

10918

207

Jul-00

10211

173

Jul-01

9710

190

Aug-00

10 567

140

Aug-01

11 434

210

Sep-00

10 741

138

Sep-01

10 114

206

Oct-00

10 161

133

Oct-01

10 344

188

Nov-00

10 784

138

Nov-01

10 538

208

Dec-00

8728

123

Dec-01

8228

180

Jan-01

10 020

118

Jan-02

9778

185

Feb-01

10 007

127

Feb-02

8146

178

Mar-01

implementation

Mar-02

25O OOO

2OO OOO

150000

100000

5O OOO

25O OOO

2OO OOO

150000

100000

5O OOO

P

£

£

. /

-a-

J

/

%

A

V

£

>

4

r

V

V

A

S

A-

î

A

V

*

£

*

*

s

*

i

4

r

*

*

5000 -g

Month

15 OOO

1O OOO

5000 -g

CM CM

Natural Gas

Water

Production

Electricty

Figure 4.12 Year-On-Year Energy Use and Production Output Data

This is a very common way of presenting data that we have seen in many factories, but it can be misleading because different values are shown on the same y-scale diagram. Apart from some swings in natural gas consumption (which, by the way, require an explanation) all other lines look nice and tidy and the first impression is that everything is under control.

Following such presentations, the energy team was unable to provide any definite and quantitative answer to the important question: 'Do we save, and if yes, how much?'

Why?

We have already said that a target cannot be a single figure, but must be an equation, because energy consumption depends on the actual production output. Correspondingly, base line consumption also cannot be represented by a single figure (like average annual consumption) but also must be expressed as a base line equation, which is calculated from the agreed base line data set.

An appropriate base line equation is given by the regression equation representing the bestfit line. The best-fit line goes through the center of the data scatter, hence represents an average energy/production relationship over the observed period of time (see Fig. 4.13). For this example, the base line equation for the year 1, for heavy fuel oil and production relationship, is expressed as follows:

where P is in production units and E in liters.

What is the meaning of this equation? For any actual production output 'P', it will give us energy consumption 'E'' that would have happened if no improvements were made. This consumption 'E'' we can then compare with actually measured consumption Em in the subsequent year, to determine if there are any savings:

If the above equation yields a negative value this means that energy performance improvements have been achieved since actual consumption is lower than one as given by the base line equation and calculated for the same production output.

Now, if we calculate energy savings in the same way month-by-month, we can sum the monthly savings and maintain cumulative sum of the savings achieved. This is shown in Table 4.5 and Table 4.6. Values of cumulative savings can be plotted in a so-called CUSUM graph (see Fig. 4.14a and Fig. 4.14b). Such a presentation makes any savings in the year-to-year data visible immediately! For instance, we can read from the graphs that cumulative increase of heavy fuel oil are around 450 m3, while electricity consumption has decreased by around 350 MWh.

1400

I 1200 S

f 400

I 200 0

Ee = 0.0843 P + 257.1

r

¡¿t

6000 8000 10 000 12 000 14 000 Production [Units]

6000 8000 10 000 12 000 14 000 Production [Units]

I 100

1 50

4000 6000 8000 10 000 12 000 14 000 Production [Units]

4000 6000 8000 10 000 12 000 14 000 Production [Units]

Figure 4.13 Baseline Equations

Table 4.5 CUSUM Data for ECC 1

Month

Production

Electricity

Predicted

Difference

CUSUM

Units

MWh

MWh

MWh

MWh

Apr-Ol

8787

848

1010

-162

-167

May-Ol

10 878

1153

1192

-39

-206

Jun-Ol

10918

1134

1196

-62

-268

Jul-Ol

9710

1059

1090

-32

-300

Aug-Ol

11 434

1227

1240

-14

-313

Sep-Ol

10 114

1124

1126

-1

-315

Oct-Ol

10 344

1126

1146

-20

-334

Nov-Ol

10 538

1135

1163

-28

-362

Dec-Ol

8228

1033

962

72

-291

Jan-O2

9778

1042

1096

-54

-345

Feb-O2

8146

962

954

8

-337

Mar-O2

The CUSUM graph has another useful feature: whenever the slope or direction of the line changes at a particular data kink, it means that something has happened in the monitored process at that instance of time. Again, the causes of change cannot be explained by the CUSUM graph. Analytical reasoning skills must be applied to analyze operational conditions and circumstances at the time of change and extract reasons for the detected change. These are the permanent tasks of performance improvement teams -to keep the performance under control and provide explanations for any detected variation, to apply corrective actions and strive for continuous improvements.

The CUSUM graph is a very good tool to make these objectives clear and their achievements immediately recognizable.

If there is a permanent change in the underlying energy/production relationship (such as the addition of a new machine), that will affect the CUSUM graph, then the base line equation must be adjusted. The CUSUM method offers a basic and effective mechanism for verification of performance improvements and monitoring and for actual savings verification at individual ECCs.

The value of CUSUM as a management tool is presented in Figure 4.15. If there are no changes in consumption patterns, the CUSUM curve will simply fluctuate around a horizontal line. However, any change which produces saving or waste will cause the CUSUM curve to change direction. It will then continue in this direction until there is another change in consumption when the curve changes direction again. The common CUSUM curve consists of a number of straight sections. Intersecting points represent the start of the impact of a new performance improvement measure or of some events that make performance worse. Extrapolating these segments to the present day shows immediately the cumulative savings (or wastes) to date from respective performance improvement measures. The advantage is that

Table 4.6 CUSUM Data for ECC 1

Month

Production

HFO

Predicted

Difference

CUSUM

Units

m3

m3

m3

m3

Apr-01

8787

181

136

20.8

20.8

May-01

10 878

202

163

42.4

63.2

Jun-01

10918

207

163

47.0

110.2

Jul-01

9710

190

148

41.6

151.8

Aug-01

11 434

210

170

51.4

203.2

Sep-01

10 114

206

153

53.3

256.5

Oct-01

10 344

188

156

43.2

299.7

Nov-01

10 538

208

158

50.6

350.3

Dec-01

8228

180

129

35.5

385.8

Jan-02

9778

185

149

36.1

421.8

Feb-02

8146

178

128

34.9

456.7

Mar-02

-200

-300

-400

-200

-300

Apr-00

May-00

Jun-00

JuI-00

Aug-00

Sep-00

00-130

Nov-00

Dec-00

Jan-01

1-0 be F

M

pr A

May-01

Jun-01

JuI-01

Aug-01

O

Nov-01

Dec-01

Jan-02

Feb-02

Mar-02

Month

Figure 4.14 CUSUM Graphs for Savings Verification

Figure 4.14 (Continued)

Figure 4.14 (Continued)

Figure 4.15 Evaluating Energy Savings Using CUSUM

Month

Figure 4.15 Evaluating Energy Savings Using CUSUM

values on CUSUM graphs are measured in physical units and can be converted simply in money terms by multiplying the measured amounts with corresponding unit price.

This example illustrates how the CUSUM technique shows the results of performance improvements. However, CUSUM analysis is useful only if the measurements are reliable, the data are adequate for the purpose of performance evaluation and the data processing is performed carefully and consistently.

4.9 Moving Toward Targets - Process of Change

The energy and environmental performance improvement process starts from a known performance level and moves toward a clear target over a given time frame. The base line or reference performance level is determined by the best-fit line and corresponding regression equation for the reference year or another selected period of time.

A reasonable target for improvements over the next 12 month period can be based on the best performance achieved during the reference year and expressed by the target line and corresponding equation based on the selected data points. For our original case study, the base line and target line are given by the following equations:

The target line can be used to evaluate progress in performance improvement and variability in energy and raw material use. It can also be used to quantify overall energy savings related to the given target. For instance, if we calculate the estimated monthly energy consumption by using the target equation and substituting P for each month from the reference period, and then take the sum of the resulting monthly Es, we will get the value of:

The difference between actual annual energy use Ej and the value Ej, gives the targeted annual amount of energy that needs to be saved:

Targeted Annual Energy SAvings Amount = Es - Ej = 18,026 - 15,479 = 2547 ton (4.7)

The equivalent emission reductions consequent to this energy performance improvement are as follows:

ACO2 = (710.37kg CO2/h - 612.19kg CO2/h) • 5000h/a = 490900 [kg/a] or 491 [t/a]

This is calculated by using Software No. 7 in Toolbox on the accompanying website.

To achieve these savings, we will need to improve energy management practice and performance so that at the end of a given time interval (a year) all the points on the scatter diagram are below or around the target line, which is based on only three data points from the reference period! This will be, of course, an ideal case.

The CUSUM method based on the base line equation from a reference year will give us the actual amount of cumulative energy savings over the observed period. However, if we want to press for more performance improvements after first 12 months of EEMS operation, we can draw a new base line for this new 12 month data set and then establish a new target line for the next period. Or alternatively, we can decide to change only the target line and to evaluate the results against the original base line from the reference year.

Every improvement process is at the same time a process of change. Operational practices need to be adjusted; energy and material use controlled more strictly; data evaluation done regularly; two-way communication maintained continuously, etc. All of these issues require a change in people's attitude, behavior and habits. And these are the most difficult aspects of change. Therefore, the process of introducing an energy and environmental management system in a factory will face similar challenges to the introduction of any other change.

We will go through the practical steps of this process in the chapter that follows.

4.10 Bibliography

Bannister, K.E. (1999) Energy Reduction through Improved Maintenance Practices, Industrial Press, Inc.

Behrens, W., Hawranek, P.M. (1991) Manual for the Preparation of Industrial Feasibility Studies, UNIDO, Vienna.

Environmental Engineer's Handbook (1999) Ed. D.H.F. Lui, B. Liptak, CRC Press. http://www.iso.org/iso/iso_14000_ essentials.

Gvozdenac, D., Morvay, Z., Kljajic, M. (2003) Energy audit of cooling towers, PSU-UNS International Conference 2003: Energy and the Environment, Prince Songkla University, Hat Yai, Thailand, 11-12 December.

Ministry of Energy, Philippines (1985) Guides to quick estimates of energy costs for industrial use.

Morvay, Z. et al. (1991) Energy Audit in Industry, published in proceedings: Environmental Management in Ukraine, held in Geneva, Switzerland (English)

Morvay, Z. et al. (1993) Energy Auditing Handbook, UNIDO, (English)

Morvay, Z., Gvozdenac, D., Kosir, M. (1999) Systematic Approach to Energy Conservation in the Food Processing Industry, Manual for Training Programme 'Energy Conservation in Industry' (training course performed in Thailand), supported by Thai ENCON Fund, EU Thermie, ENEP program.

Schnell, K.B., Brown, C.A. (2002) Air Pollution Technology Handbook, CRC Press.

The Engineering Handbook (1998) Editor-in-Chief, R. C. Dorf, CRC Press.

UNEP (1990) Environmental Auditing, UNEP, Technical report series No. 2.

UNEP (1991) Audit and reduction manual for Industrial emissions and wastes, UNEP, Technical report series No. 7.

Welch, T.E. (1998) Moving Beyond Environmental Compliance: A Handbook for Integrating Pollution Prevention with ISO 14000, CRC-Press; February 1.

Welsh Office (1989) Environmental Assessment - A guide to procedures, Department of Environment, , HMSO, London.

Welsh Office (1990) Integrated Pollution Control - A Practical Guide, Department of Environment, Welsh Office, HMSO, London.

Whitelow, K. (2004) ISO 14001 Environmental Systems Handbook, Second Edition, Butterworth-Heinemann; October 19.

Was this article helpful?

0 0

Post a comment