The origins of modern financial accounting standards are centuries old. The origins of greenhouse gas (GHG) accounting, by contrast are two or three decades old. Is it any wonder then that we’re still working out the details? At Scope 5, much of our work revolves around helping our clients develop methodologies for and conduct, their GHG inventories. We live in the world of GHG accounting.
For most of our clients, GHG accounting is still voluntary. Accordingly, they’ve been concerned with getting the numbers defensibly ‘in the ballpark’. But more and more clients are seeing the need to apply stricter standards than they have in the past. They’re facing audits and carbon fees and material errors in their GHG inventories are suddenly consequential.
As we work with these clients to firm up their methodologies, one of the issues that we address concerns the manner in which emissions should be distributed over time.
Before delving into details, it’s important to note that:
- Most emissions arise from an underlying activity and in most cases, deciding how to distribute resulting emissions over time amounts to deciding how to distribute the underlying activity over time.
- Different methods apply, depending on the nature of the activity.
- Depending on the underlying activity, applying one method or another could have material consequences. In other cases, the difference between applying one method or another might be minimal or inconsequential.
In this post, we’ll look at one specific example – prorating utility related emissions to a specific calendar year. Many of our clients are required to produce an emissions inventory that accounts for their emissions over a calendar year. A material part of their emissions results from underlying utility activity.
An important part of quantifying these utility related emissions entails coming up with emissions factors – the numbers that convert from activity units (such as kWhs of electricity) to emissions (tonnes of CO). But emissions factors are the topic of another post. Here, we’re concerned with the distribution of the emissions over time.
Let’s consider a hypothetical enterprise – Acme Corp. The sustainability team at Acme is working on their 2015 inventory. They’ve collected a stack of electricity bills that quantifies their electricity activity for calendar year 2015. It should be a simple matter to total the usage for the stack and multiply by the appropriate emissions factor to get the resulting electricity emissions for the year.
Looking at the stack of bills, we see 12 bills. Each bill has a few dates on it. The bill at the top of the stack has a bill date of 1/14/2015. We flip through the stack and see that each bill has a bill date for the next month, in sequence, with the last in the stack, dated in December 2015. Perfect!
But wait – what does this bill date mean? It’s the date the bill was generated. That date has little to do with when the related emissions were emitted. Looking more closely, we see that each bill also has a service start date and a service end date. Now these are more meaningful.
The bill at the top of the stack has a service start date of ‘12/4/2014’ and a service end date of ‘1/5/2015’. It shows a usage of 413,378 kWh. What’s it really telling us? It’s telling us that, over the 33 days spanning January 4th, 2014 to January 5th, 2015 Acme used 413,378 kWh of electricity. This may be the ‘January 2015 bill’ but 28 of the 33 days over which it quantifies electricity use were actually in 2014, not in 2015! It would be an error to include 413,378 kWh of electricity and the resulting emissions in Acme’s 2015 emissions inventory.
How bad can this get?
Such an error might not be that meaningful – after all the activity for December 2015 would be omitted since we don’t have the January 2016 bill in our stack. So – we’d include a little extra from 2014, but we’d exclude a little from 2015 – no big deal…
But what if December 2014 was an anomaly – imagine that Acme heats its warehouses using electricity and December 2014 was a particularly cold month. Digging up the January 2016 bill we see that the omitted activity for December 2015 was only 84,089 kWh – December 2015 was an unusually warm month. This scenario is illustrated in the following figure:
The figure shows two columns. The first column represents a stack of thirteen bills and the corresponding service periods. Twelve of these have a bill date in calendar year 2015. The thirteenth has a bill date in January 2016. The service periods corresponding to the first and last bill actually span 2 years. The first bill’s service period spans part of December 2014 and part of January 2015. The last bill’s service period spans part of December 2015 and part of January 2016. If we total the activity in the first twelve bills, we include some activity that occurred in 2014 and we omit some activity that occurred in 2015. If we total the last twelve bills, we include some activity that occurred in 2016 and we omit some activity that occurred in 2015. So – we need to prorate the first and last bill in order to correctly sum all activity attributable to 2015 and none of the activity attributable to 2016. In order to do so, we’re going to assume that the activity for any service period is uniformly distributed across the days in that service period.
The second column in the figure above represents calendar year 2015. It’s vertically aligned with the first column to correspond to the overlaps of the first and last service periods with the calendar year. Specifically - the first service period spans a total of 33 days but only 5 are in calendar year 2015. So – the top of the calendar year column starts 5/33 of the way above the end of the first service period. The January 2015 bill contributes only
5/33 X 412,378 kWh = 62,633 kWh
to calendar year 2015.
The service period of the January 2016 bill spans 31 days, of which 26 are in 2015. The January 2016 bill contributes
26/31 x 84,089 kWh = 70,526 kWh
to calendar year 2015.
The correct prorated total is illustrated in the second column, which sums the:
- prorated amount for January 2015 (from the January 2015 bill)
- total amounts from the February 2015 bill through the December 2015 bill
- prorated amount for December 2015 (from the January 2016 bill)
The eleven bills starting with the February 2015 bill and ending with the December 2015 bill total 1,219,421 kWh. This total represents activity that is included entirely within calendar year 2015. Adding the prorated amounts attributable to 2015 from the January 2015 bill and the January 2016 bill give us the prorated total for 2015 of 1,352,580 kWh.
If we had simply added the activity for the twelve 2015 bills, we would have arrived at 1,632,799 kWh. This is 280,219 kWh more than the prorated amount. Without prorating, we would be overstating our 2015 electricity activity (and resulting emissions) by more than 20%!
The example illustrated is a bit extreme – the January 2015 bill includes almost five times as much activity as the January 2016 bill – but the point remains valid: failure to prorate can introduce material errors when calculating GHG emissions.
In summary, when calculating GHG emissions for a specific time period, it’s important to carefully consider how the underlying activity aligns with the time period of interest. In the case of utility activity, numbers are often taken from bills that reflect meter readings corresponding to particular service periods. In this case, the activity for service periods that are not entirely included within the period of interest must be prorated. Often, the best we can do is to prorate assuming uniform distribution of activity over the corresponding service period.
Comments