Energy Measurement In Process Industries – Why You Are Probably Measuring The Wrong Thing

If you are a typical company in the process industry then you are probably measuring your attempts at energy management by measuring your Specific Energy Consumption (SEC) on a monthly basis.

Depending on the process this may be in ‘kWh/kg' or some similar metric. You then use this metric to see if you are getting more ‘energy efficient’ or not, i.e., if it goes down then you feel good and if it goes up then you feel bad.

However, SEC is a fundamentally flawed metric of energy efficiency for process industries. It is not only the wrong metric for monthly tracking but can also lead you seriously astray. This may seem heresy when so many people use it but the proof of this is simple mathematics.

Do not use SEC as a metric for monthly energy performance or for comparison between sites.

It is a fundamentally flawed metric for these purposes.

Energy use can be related to either ‘activity’ or ‘condition’ drivers.

• An ‘activity’ driver is where the energy use varies with an internal activity, such as production volume.
• A ‘condition’ driver is where the energy use varies with an external influence, such as the weather.

For a site with a single main process that uses electricity or gas, then energy use is directly related to the value of the activity driver. This is shown in Figure 1 and is the normal case for many process industries which use only a single process.

This shows that the site has a total energy use made up of:

• The base load – this is the fixed energy use from lights, air leaks and other energy uses that are constant over the month.
• The process (variable) – this is the load that varies with production volume, i.e., process more material and you use more energy.

The total site load varies with production volume according to:

kWh = A x Production volume + B

and this equation is the Performance Characteristic Line (PCL).

The PCL shows that the total site energy use varies directly with the amount of material processed and nobody should be surprised by this. You can even test the relationship for yourself by plotting your own monthly production volume versus energy use.

We have measured data from a variety of process industries ranging from plastics processing (over 500 sites) through to cold rolling and heat treatment of steel and they all show a relationship between the monthly production volume and the monthly energy use as shown in Figure 1.

So, what is the SEC and how does it relate to this graph? In terms of the PCL for a typical site, we can visualise the individual monthly measurement of SEC (kWh/kg) as the slope of the line drawn between the origin and an individual monthly data point. The slope of the line for each month will be the monthly SEC value (see Figure 2).

It is immediately obvious that the simple monthly SEC type of measurement is affected by the production volume. Two monthly results located on exactly the same PCL and representing the exactly the same basic energy performance will have highly different SEC measurements if the monthly production volume is different.

Increasing the production volume will therefore automatically decrease the SEC because the base load will be amortised over a greater process load and lead to the impression that energy efficiency is improving.

Decreasing the production volume will also automatically increase the SEC because the base load will be amortised over a smaller process load and lead to the impression that energy efficiency is getting worse.

This is simple amortisation of the base load over the process load, i.e., high production volume = low kWh/kg and low production volume = high kWh/kg. There is no mystery here, it is simple mathematics. This is the same as calculating the unit cost of a product when there are production volume changes.

A company will have fixed and variable costs and increasing production will amortise the fixed cost over a larger variable cost and the unit cost of a product will appear to decrease. Every accountant should be familiar with this concept and would never think of punishing the production manager based on variations in a volume related unit cost.

Let’s take a look at some real data from a plastics processor. Figure 3 shows the production volume and monthly SEC for this site.

The production volume varies with the month but the overall trend is for increased monthly production volume. The SEC (in kWh/kg) also varies with the month but the overall trend is downwards and the energy efficiency is apparently increasing, i.e., it apparently takes less kWh to produce a kilogramme of finished product.

Unfortunately, all is not as it seems.

What is the connection between the two lines? It is obvious that the SEC is in fact the mirror image (ignore the scaling) and that when production volume is high then the SEC is low and vice versa. To verify this, we can plot the SEC versus the production volume to see the exact relationship, this is shown in Figure 4:

For this site, the SEC can vary between 1.37 and 2.20 simply through changes in production volume. Monitoring the SEC as a metric of energy efficiency is a worthless exercise.

Sites must therefore be careful in monitoring energy efficiency changes by simply comparing SEC values; these can be affected by simple changes in production volume rather than real changes in the site energy efficiency. This will be less significant where the base load is low in comparison to the process load but the simple number is almost always misleading.

This is not a problem when production volumes are rising and the management team sees a continuously decreasing SEC. They are happy to accept the congratulations for doing nothing at all. When production volumes are decreasing and the SEC is increasing despite any efforts they may be making then they are less happy to accept the criticism.

If you are measuring your energy performance by measuring the monthly SEC then you may have noticed that the ‘kWh/kg' value goes up and down quite erratically even when you have changed nothing.

As an example of what can happen, a colleague was using monthly SEC as his performance metric and was frustrated by the seemingly random variations in SEC each month despite carrying out good work in energy management.

When the reasons were explained he was overjoyed and found that his PCL was really excellent and actually showed the real progress he was making.

Use the monthly SEC value as a metric only if you are not capable of holding more than one number in your head at a time.

What are the alternatives?

Many companies use the monthly SEC as their fundamental metric for energy efficiency but as a metric for progress it is worse than useless because it can reward bad behaviour and punish good behaviour – it all depends on which way the production volume is moving and the size of the base load for the site.

The most appropriate metric for assessment of energy performance is a site-based PCL as shown in Figure 1. This takes into account the specific nature of every site, i.e., the base load and the process load and automatically corrects for production volume variations.

Despite the simple mathematics which show that SEC is a flawed metric for monthly performance, there are many companies who are determined, against all arguments, to use the monthly SEC as a performance metric.

This is perhaps because their managers are not capable of holding more than one number in their head at a time – a sad indictment of the quality of the managers. If this is the case, then it is possible to use a moving average of the SEC over a period of at least 9 months to smooth out production volume variations.

If production volumes are strongly increasing then even this method will not eliminate the effect of base load amortisation but it is marginally better than using a simple monthly SEC.

When used internally as a monthly performance metric, SEC is flawed but this is only internal and companies are always free to make their own decisions. It is even worse when governments use the same flawed metric as a measure of industry performance and reward or punish sectors on the basis of increasing or decreasing SEC.

Summary

If you are going to use a performance metric then the first action should be to check that the metric actually measures what you want to improve. In the case of SEC, the metric measures the wrong thing and can does not help you to get better. If you are using it then think again and start to measure something that really helps you to get better.