MassCEC has extended the deadline for our whole-home air-source heat pump pilot through December 18, 2020 or as long as there is available budget. The pilot launched in May 2019, and so far we have awarded a total of $200,000 to 61 projects out of an available budget of $500,000. Of those 61 projects, 34 are existing building retrofits and 27 are new construction or gut rehab projects. About a third of the projects have received an income-based or affordable housing adder. See our previous monthly blog updates for more information about the pilot.
For the June pilot update, I’d like to focus on what MassCEC has been learning about estimating heating load from natural gas usage.
Goals & Objectives of Reviewing Natural Gas Usage
In the March blog, I announced that we had started to require 36 months (3 years) of gas usage data for existing building heat pump retrofit projects. My colleague Jacqueline Guyol put together instructions on how customers can download 3 years of gas bill data from Columbia Gas, Eversource, or National Grid. Our goal in requesting the gas usage data was to compare it to the installer’s Manual J heat load calculation to see if we could gain any insights into estimating heating loads. Specifically, I’ll admit that I was hoping that past natural gas usage could be used to accurately calculate a home’s heating load, since we have continued to see challenges and limitations with the Manual J heat load calculations, some of which I discuss in the blog from last September. After talking through the methodology with our technical consultant Bruce Harley and seeing data from a few example projects, I have a better appreciation for the limitations of natural gas usage data (for reasons I’ll discuss in the examples below), but I still think it can be a useful check on the home’s heating load.
Methodology for Estimating Heat Load from Natural Gas Usage
Our technical consultant Bruce Harley developed a methodology for doing a simple estimate of the peak heating load based on past natural gas usage. Based on his methodology, my colleague Jacqueline Guyol created a template for taking applicant’s past natural gas usage data and generating an estimated peak heating load. This estimate includes a step for scaling the observed natural gas usage up or down based on how cold the winter was compared to a “normal” winter. All three of the past winters have been warmer than typical meteorological year data, so that means that normalizing the data increases the heating load from what recent gas bill usage shows. Of course, as the climate warms, it’s possible that the data we are using for the winter design temperature is too cold, but for now we’re still making this adjustment.
If you’re interested in the details of how we’re estimating peak heating load from natural gas bill data, check out this hypothetical example. I want to emphasize that this methodology is a rough estimate. If you’re interested in learning talking to us more about how we’re doing these estimates, email me and Jacqueline.
For the projects where we have analyzed the natural gas bill data so far, we have had a variety of outcomes. For the three examples below, we found the installer’s Manual J calculation to be the same (Example 1), higher (Example 2), and lower (Example 3) than the heat load suggested by the natural gas usage data. Example 3 in particular highlights some of the limitations of using past natural gas usage to calculate a home’s heating load.
- Example 1: Manual J and natural gas usage about the same. We received an application for a 527 square foot cottage in Cape Cod where the installer estimated a design heating load of 11,726 btu/hr. Looking at the customer’s past natural gas usage and using the methodology described above, we came to a design heating load of 13,945 btu/hr. In this case, the natural gas usage data suggests that the installer’s Manual J is a good value to use for the sizing the equipment.
- Example 2: Manual J high than natural gas usage. We got another application for a 2600 square foot home in Brookline where the installer estimated a Manual J design heating load of 76,840 btu/hr. Using our methodology for estimating peak heat load from the gas bill data, the home’s peak heating load would be 62,637 btu/hr, which suggests that the installer’s Manual J may be overstating the home’s heating load. Interestingly, the installer proposed a system that is more in line with the heat load suggested by the natural gas billing data.
- Example 3: Manual J lower than natural gas usage. We got an application for an 822 square foot condo on the third floor of a triple decker in Somerville. The installer’s Manual J calculation estimated the heating load as 34,749 btu/hr, which looked really high to us. However, when we looked at the natural gas usage for the home, with our standard assumptions the estimated heat load would have been 91,377 btu/hr! However, our standard assumption for heating system efficiency was 80%, while for this home the natural gas steam boiler was over 20 years old and the installer estimated that the boiler was only 50%-75% efficient. Since the unit is on the third floor the overall system may be even less efficient as uninsulated pipes lose heat on the way up the third floor. Also, the previous tenants were older and kept the heat very high. This example shows that just looking at a rough estimate based on the natural gas usage could lead an installer to overestimate the home’s heating load. In this case, the installer sized the heat pump to the Manual J design load. The owner is planning to take advantage of Mass Save’s limited time offer for 100% off approved insulation to do some weatherization for the unit. So even with the high natural gas bills, it seems likely that if anything the heat pumps will be oversized.
Working through these examples has shown me that even if a customer has three years of natural gas bills, that’s not a silver bullet for understanding the home’s heat load. However, I do think that the natural gas bill data is a useful check for the Manual J heat load. As we get more projects with this data, we will analyze the natural gas usage for any other lessons learned or trends and share those in future months.
If you are a homeowner or installer working on a potential whole-home air-source heat pump project (or anyone else!) please email me if you have any questions or feedback.