How to Set an Effective Energy Baseline and Performance Indicators

An energy baseline and clear performance indicators help an organisation prove whether its energy performance has improved. If these are not set correctly, it becomes difficult to demonstrate actual improvement.

Step 1: Identify What Affects Energy Consumption

Start by identifying the factors that regularly influence how much energy your organisation uses. These are called relevant variables.

Common examples include:

• Production volume

• Product mix

• Operating hours

• Ambient temperature or degree days

• Building occupancy

• IT load in data centres

Also identify static factors, such as building size, installed equipment and process design. These normally do not change frequently, but they become important when there is a major change, because the energy baseline may then need to be updated.

Step 2: Collect Data at the Right Level of Detail

Monthly electricity or utility bills can provide a basic picture of overall energy use, but they are usually not detailed enough for proper energy analysis.

For significant energy uses, sub-metering is useful. Half-hourly energy data at site level is a good starting point because it can show:

• Base energy consumption

• Load patterns

• Shift patterns

• Changes in consumption throughout the day

Collect data for the relevant variables over the same periods. For example, production and weather data should match the same dates as your energy consumption data. Mismatched periods can lead to inaccurate results.

Step 3: Select the Baseline Period

Choose a period that represents normal operating conditions.

A 12-month period is commonly used because it captures seasonal changes. Avoid periods that include unusual events such as:

• Major shutdowns

• Plant expansion

• Unusually high or low production

• Other abnormal operating conditions

Document why the selected period was chosen. If no period perfectly represents normal operations, clearly explain the reason for your selection.

Step 4: Understand the Relationship Between Energy Use and Operations

Do not look only at total energy consumption. Study how energy consumption changes when the main operating factors change.

For many manufacturing organisations, production volume is the main factor. Energy consumption often increases as production increases.

The relationship may also show a baseline level of energy use even when production is very low or zero. This is known as baseload consumption and can include energy used for essential equipment, lighting, security or other continuous operations.

If several factors influence energy consumption, multiple regression can be used to understand their combined effect. This can also be done using a spreadsheet.

Step 5: Create Meaningful Energy Performance Indicators

A simple energy performance indicator can be:

• Energy per tonne produced

• Energy per unit produced

• Energy per square metre

However, simple ratios have limitations.

For example, energy per unit may appear worse when production falls because some energy is still being used even when production is low. This does not necessarily mean that energy efficiency has declined.

Where this is an issue, a regression model can provide a better measure. Compare the actual energy consumed with the amount of energy the model predicts for the same operating conditions. The difference helps show whether performance has improved or declined.

Create indicators for each significant energy use, rather than relying only on one indicator for the entire site. Site-level figures can hide improvements in one system and increases in another.

Step 6: Define the Baseline Adjustment Rules in Advance

The organisation should establish clear rules for when and how the energy baseline will be adjusted.

Adjustment may be required when:

• The existing indicators no longer properly represent energy use

• There are major changes to static factors

• A predefined adjustment method requires it

Document the adjustment method before it is needed. Changing the baseline only after seeing an unfavourable result can raise questions during an audit.

Step 7: Validate the Model

Test the model using data from a period that was not used to create the original model.

If the model predicts energy consumption reasonably well for this separate period, it is more likely to be reliable.

If the prediction is significantly different from actual consumption, check whether an important variable has been missed.

Common Mistakes to Avoid

1. Using total energy consumption as the only indicator without normalising the data.

2. Not identifying the relevant variables that affect energy consumption.

3. Selecting a baseline period simply because data is available rather than because it represents normal operations.

4. Using simple energy ratios when significant baseload consumption makes them misleading.

5. Measuring performance only at site level instead of for significant energy uses.

6. Adjusting the baseline without having a predefined adjustment method.

7. Using energy consumption data and relevant variable data from different time periods.

8. Not validating the model against an independent period.

Key Takeaway

A good ISO 50001 energy baseline is more than a historical energy consumption figure. It should show the relationship between energy use and the factors that drive consumption.

When the baseline, relevant variables and performance indicators are properly defined, an organisation can more clearly demonstrate whether its energy performance has genuinely improved.