Estimating missing activity data  

The best estimation methodologies are based on the highest quality data sources that either correlate strongly or otherwise predict the estimated data. That said, there is a preferred order for these methods based on the precision of the approach:  

Method 1: Repair incomplete raw data  

Before attempting to fully estimate emissions for a given activity, review available data to see if it can be repaired by estimating any missing elements. This section describes three recommended repair approaches applied to example situations where only a portion of a leased office space’s electric utility bills are available:  

To estimate the missing consumption data from 2/26/2021 – 3/23/2021, apply the interpolation method:  

Example 2:  

To estimate the missing consumption data from 11/6/2021 – 12/4/2021 and 12/5/2021 – 12/31/2021, apply the extrapolation method:  

Example 3:  

To estimate the missing consumption data from 1/1/2021 – 1/28/2021, apply the substitution method:  

Method 2: Develop internal intensities  

Sometimes the repair methods above will be unavailable or impractical. Maybe there is a fuel tracking system that has only been installed for a portion of a fleet, or utility data is only available for some leased office spaces. In these scenarios, it may be possible to estimate consumption by creating an internal energy usage intensity (EUI) metric (e.g., kWh electricity per square foot, or miles per vehicle) and applying that intensity to the missing portion of the data.   

For example, a common situation faced by building occupants without sub-metered utilities is that energy consumption is only tracked at the whole-building level. In this case, the whole building data can be allocated to the leased portion of the building either by area (e.g., for electricity, which is tied to the space) or by employee (e.g., for water, which is tied to employee use).  

To estimate Office D’s missing consumption data, use the square footage and electricity consumption from similar properties to calculate the internal EUI:  

Method 3: Utilize correlated data  

When there aren’t enough data points to develop an internal EUI to estimate missing consumption data, try to identify an adjacent available data point that correlates to the missing data. For instance, your company may maintain an accounting system that tracks utility expenses, which can be used to fill data gaps if electricity consumption volumes are unavailable. As most utility costs scale with the volume of fuel or electricity consumed, spending can be used to approximate energy data by applying a utility rate (e.g., USD/kWh). Be sure to exclude taxes/VAT and service fees and use rates from your data gap year to avoid inflation adjustments.  

Example 5:  

To estimate the electricity consumption of offices where expense data is available, but consumption data is not:  

Method 4: Apply industry average intensities  

If raw datasets prove to be unavailable or unusable, it may be necessary to find and apply an appropriate industry average EUI. These datasets are typically developed from a national or regional sample survey that collects information on commercial buildings, including their energy-related building characteristics and energy usage data. Some of the most well-known sources include:  

While this approach may be appealing for its simplicity, it doesn’t allow for accurate tracking of trends over time or realize the results of your company’s emissions reduction efforts (i.e., energy efficiency projects, behavioral changes, etc.). Also, this method isn’t suitable for buildings with non-standard activities, such as manufacturing plants. Company GHG inventories that utilize an industry average in their base year should recalculate their base year when more accurate consumption data becomes available.  

Example 6:  

Office A is a 2,000 square foot leased office space in Arizona in the United States. Due to the structure of the lease agreement, electricity consumption data is not available for this office.