Navigation:  Introduction - Energy Management in Buildings > Energy Accounting >

Baseline Model

Previous page Next page
  rev. 2011-06-14        

For additional information on the Baseline Model, please see our video in the Energy Basics series, available on our website.

A Baseline Model is a mathematical formula that captures a snapshot of building performance during a past period known as the Baseline Period. Once defined, the Baseline Model allows us to extrapolate building performance from the Baseline Period to current conditions. This allows for a direct comparison of past and present performance in which we can measure with certainty any change in building performance independently of the influence of weather and other variables which can make such comparison very difficult.  In ManagingEnergy it is a mathematical formula derived according to industry standards.  The most commonly used independent variables are billing period (number of days), heating degree days (a measure of heating load), and cooling degree days (a measure of cooling load).  ManagingEnergy also allows inclusion of additional user-defined independent Facility Variables such as production volume, occupied suites, or meals served.

When it comes time to measure the performance of energy conservation efforts, to find out how much is actually being saved, the Baseline Model can be thought of as an imaginary, mathematically-defined What Would Have Happened utility meter.  The savings are the difference between the actual meter and this mathematical meter.

In the Baseline Model, consumption is a linear function of time and outside temperature.**  Baseline consumption increases in a predictable way as time passes and heating and/or cooling load arises.  With a formula, the analyst can take the time and outside temperature conditions for any billing period, plug those values into the Baseline Model, and find out exactly what the baseline consumption would have been under the prevailing conditions.  That will provide a valid comparison even if the current conditions are quite different from those during the equivalent baseline period.  The consumption result produced by the Baseline Model can be compared directly against the current invoice.


                                                                 A Baseline Model is used  to factor out the differences (between the baseline period and the comparison period) in independent variables like weather differences, billing period differences, and production volumes.  It provides a basis for valid apples to apples comparison.


To build a Baseline Formula, an analyst needs to know three things for each invoice during the Baseline Period:

The consumption, which is available directly from the invoice
The duration of each billing period, which is also available from the invoice
The outside temperature conditions during the billing period, available from a weather service

Each invoice is a single sample. With a reasonable number of sample points, the energy analyst can use linear regression (a standard statistical method) to define the baseline formula.  Note that there is no mandatory number of baseline invoices required.  ManagingEnergy can build a baseline with as few as six, but more than six means greater statistical certainty.


                                                                 Even though ManagingEnergy allows it, don't be tempted to use all available invoices to create the baseline.  You want to use only entries that reflect current facility conditions.  Buildings and occupancy change over time, so more recent values are likely to be more relevant.  In most cases, 12 to 24 recent invoices provide a good sample size for a baseline.

Form of the Baseline Model

A typical Baseline Model looks something like this:

Baseline Model for a heating-sensitive natural gas meter

Baseline Model for a heating-sensitive natural gas meter

This example is for a natural gas meter.  The baseline consumption of natural gas, in Mcf (thousands of cu.ft.), is the sum of two parts:

1.A base which depends on the number of days.
2.A heating portion which depends on the number of Heating Degree Days, a measure of the coldness over the period.  In this case, the Heating Degree Days are calculated based on a Heating Balance Point of 20 Celsius, which means that heating is expected whenever the outside temperature is below that value.  ManagingEnergy automatically figures out the best Heating Balance Point as well as the two multipliers shown in the model.


A second example, in which an electrical meter is sensitive to both heating and cooling load, is shown below.  This would be the case where a building has electric heat operating in winter and air conditioning in the summer.  In this example, the Heating Balance Point and the Cooling Balance Point are both 18 Celsius.  They will not necessarily be the same.  In fact there is usually a gap of a few degrees between the heating and cooling balance points, reflecting an outside temperature range where neither heating nor cooling is required.

Baseline Model for an electrical meter that is sensitive to both heating and cooling load

Baseline Model for an electrical meter that is sensitive to both heating and cooling load


A third example would be an electrical meter that feeds only lighting.  Because there is no heating or cooling equipment connected, consumption for the meter is not weather-sensitive and depends only on the length of the period.

Baseline Model for an electrical meter that is non weather-sensitive

Baseline Model for an electrical meter that is non weather-sensitive


Missing or Bad Baseline Readings

When you think of meter readings as data samples, you'll realize that the Baseline Model process can help you out of awkward data problems.  If you are missing a particular invoice during the baseline period, it's not a big problem  because a valid baseline formula can still be produced with the surrounding readings.  You're simply working with one less sample.  Similarly you can intentionally leave out atypical readings from the baseline, where consumption is exceptionally high or low.  These could be caused, for example, by a facility shutdown or a major operational problem that is not expected to be repeated.


                                    The Baseline Period can be the 12 consecutive invoices, but it doesn't have to be.  ManagingEnergy can produce a baseline with as few as six meter readings.


Use the Auto Baseline Model function have have ManagingEnergy automatically creates Baseline Models.


Adjusted Baselines

The adjusted baseline represents the baseline energy use pattern but adjusted to the conditions of the reporting period. In other words, the adjusted baseline represents the energy use which would have occurred if the building were performing as it was during the baseline period.

Avoided Energy Use (or Savings) = Adjusted-Baseline Energy - Reporting-Period Energy ± Non-Routine Adjustments of baseline Energy to reporting period conditions

ManagingEnergy adjusts the baseline automatically for routine adjustments (weather, occupancy, production, for example) and non-routine adjustments (Energy Conservation Measures (ECM), static factor changes, etc.)

Opportunities (our term for ECM) will only ‘adjust’ the baseline when they are marked as complete and the reporting period is after the implementation date of the opportunity.

If a client does not have access to the advanced analysis tab they do not have model adjustments in the system and hence the Adjusted Baseline = Baseline - 0 = Baseline.

For a more detailed example, see our FAQ entry Why Does my Adjusted Baseline equal my Baseline?


** In fact, the baseline formula can include additional independent variables in addition to time and outside temperature.  That possibility is covered elsewhere in this manual.


Related Topics

Baselines and Usage Profiles

Energy Accounting

Energy Accounting Challenges


intro_Baseline_Model         ©2014 Managing Energy Inc.