Energy improvement: look at the big picture for real savings

Paul Boughton
There is more to saving energy than fixing steam and compressed air leaks, investing in energy efficient motors and having boiler efficiency surveys. After all, if reducing energy consumption were industry's only concern, then the simple solution would be to turn everything off!

Obviously that is not an option, so in order to make savings it is important first of all to investigate exactly how we use our energy supply.

Many large companies will already employ an energy manager, but to achieve maximum energy efficiency, a holistic approach should be taken with representatives from all departments involved.

Different departments all impact on a business' energy costs. For example the purchasing department will determine the price paid for energy; the production department for how it is used; engineering for how it is distributed and technically used; finance to assign financial resources to achieve energy returns from investments; and so on.

Once a suitable team has been gathered together, the next step is to start to define and understand where and how your site consumes energy. This is best achieved by mapping out your energy distribution systems and what is connected to them.

During this process, it is important to remember that energy distribution systems themselves consume energy.


From the mapping stage, you will have identified points of energy consumption, but now you need to begin to quantify the amounts. How and why is the energy consumed? What is the pattern of energy usage at these points? This data should be delivered by an onsite metering system, which will ideally measure at 10 or five minute increments.

Add in other relevant data: what are the shift patterns you use, when are production lines or units manned up and running and when are they normally idle, what is overall equipment effectiveness like (OEE), what is the production plan and sales forecast like.

All these, and more, are relevant to a holistic energy improvement. Remember, it is not how much energy you use, but what use you make of it and all these things impact on how you use energy.

Data analysis

Once you have gathered your data, the next step is to analyse it. Energy is often mistakenly considered as an 'overhead' by organisations.

If we take the time to look at patterns in energy consumption, it soon becomes clear that this is rarely, if ever, the case. Start to look for relationships between consumption and some other variable.

For example, if your factory is heated, try relating energy consumption (electricity, gas, or steam) to degree days per week or even degree days per day (degree days are a way of measuring heating requirement against the outside weather).

This is done by plotting consumption against degree days and askingidentifing a relationship. Inspect the plot for any outliers in the data which may legitimately be removed because they relate to special circumstances such as a shut down, short running, or even a major problem on a piece of kit which significantly impacted consumption.

This will improve the level of confidence and the model of the steady state energy consumption.

Other variables to look at include occupation (are people present in the factory?), output, and any others that might occur. However your mapping exercise will highlight the relevant variables to you.

In a factory similar to a pharmaceutical site, I would expect steam consumption to be related to weather (where steam is used primarily for HVAC) and electricity to be related to occupation, ie if a factory is occupied and running, then a fixed amount of electricity is likely to be used.

Also, consider smaller areas than the gross site energy consumption, as here the relationship to the variable will likely be clearer.

With this modelling, and with the mapping, you and your team will now have a 'current state' map of your energy consumption.

Target waste

Along the way you will have learnt a lot more about how and why energy is consumed at your site, and you will have the knowledge to begin changing it such that the energy consumption and/or cost of energy per unit of product is reduced. You can target areas of waste energy use to reduce consumption, and you can generate a future state map of how energy should be used on site.

All businesses have different energy needs, but by way of example, consider the following:

Example 1. A pharmaceutical plant with a large production hall housing several different production lines but where all of the HVAC equipment is common.

Most of the production units run for five days per week, but one runs 24/7. This necessitates all of the HVAC systems running 24/7, most of the time purely for one production unit. The energy cost is therefore enormous, and wasted.

The future state in this case would be to re-arrange the production facility and HVAC system through zoning and partitioning so that the HVAC system controls can be matched to production requirements.

Example 2. Many chemical plants these days have legacy infrastructures, with steam services running to and through facilities that are no longer used, not used as much as others (similar to the pharmaceutical example), or even in some cases are no longer there.

The heat loss, and hence the energy consumption, in running steam through pipework unnecessarily is surprisingly costly.

By working through a current state map and then creating a future state map, the steam infrastructure can be modified through the use of valves to zone off portions so that steam is only fed to areas that require it.

Example 3. Centralised steam generating systems are common (and very wasteful compared with new technologies).

Understanding where energy is consumed can quickly lead to energy saving improvements. I have worked on a site where steam consumption could be directly related to the weather (degree days).

Through observation the lagging was in a poor state, and for some very extensive runs completely missing. Through a steam trap survey it was identified that a large number of steam traps were passing, also adding to the overhead cost of running the steam systems.

Where steam consumption does not relate to weather but is production dependant, the cost of running the steam infrastructure can still be measured by looking over a period such as a shut down, or a period of nil or low steam demand for whatever reason, and measuring the steam demand for that period (it is easier to measure steam demand through boiler fuel consumption than actual steam usage - steam meters are notoriously hard to use, expensive, and difficult to install).

Also, do not discount the use of theoretical information for things such as heat losses from pipework, available from sources such as the UK's Carbon Trust.

Example 4. Energy bought during the day is different from energy bought at night. And energy bought at short notice is different from energy bought ahead of time. This is a very complex area, but for the purposes of this article, suffice to say that the purchasing strategy used for energy (electricity and gas supplies) is capable of having a significant impact on energy cost.

Add into the mix other variables such as when particular operations are carried out and energy could be preferentially used when it is cheaper. This could result in an energy cost saving of 10 or 20 per cent, depending on your contract.

Example 5. Never underestimate the value of your personnel. They are your best asset, but only if you use them.

From your future state map you may identify a number of target areas: from your measurement of energy for a specific process you may find that the base load for it when not operating could be as much as 70 or 80 per cent of its running load. This may be due to pumps left running, agitators stirring, ventilation running, etc.

Putting together a small team of operators to look at ways of reducing base load consumption, with the right management support and data, will always generate ways of reducing the base load. And if it is their idea they are more likely to follow any procedures for turning off equipment to reduce consumption than if it is imposed upon them.

Poor OEE (Overall Equipment Effectiveness) will always impact on energy consumption per product unit. In batch chemical production the performance aspect of OEE is typically around the 60 per cent mark when measured against the per cent longer than they should have done, consuming energy in doing so: vessel left mixing too long, refluxed for an extra hour or two, and so on (the reasons why are myriad but all too common).Poor quality and re-worked product is the worst culprit. Again, teams of operators are very good at tackling these issues, and Picme's Masterclass product has an excellent track record in this area.

The final stage of holistic energy improvement is Control. With the modelling of the site consumption done earlier it is possible to now compare actual consumption to the modelled consumption and differences to be tracked.

A Cumulative Sum graph (CUSUM) is a good way of highlighting and amplifying differences between the model or benchmark and actual.

As energy consumption improvements were achieved, this translated into a downward sloping line indicating energy savings.

The effectiveness or efficiency of steam use has improved, and the overhead cost of running the steam systems has reduced.

Any consumptions greater than the model predicts indicate a worsening energy pattern, whereas any consumptions less than the model predicts indicate an improving energy pattern.

If your site has a key variable that indicates likely energy consumption, eg the weather, then you will now get clarity of what is going on with energy without the clutter of the unknown.

You may be over budget for energy in a particular month, but if the weather was cold you can now explain why. If your energy saving activities are bearing fruit you can even claim a hefty saving while simultaneously being over budget!

If you add in energy cost per unit into your model you can even isolate the impact of unit price movements against those budgeted for. This will give variance reporting on energy similar to accounts use for management accounts.

Improvements in consumption

Finally, in the team you set up in the beginning, you have a ready-made energy steering group to continue monitoring and driving energy consumption improvements.

Collectively they can identify and commission energy projects and initiatives taking into account all the different aspects of energy improvement work: Holistic energy improvement.

Russell Page is Senior Improvement Consultant at Process Industries Centre for Manufacturing Excellence (PICME) Limited, Manchester, UK.