Degree Day Adjustments for Heating Energy Calculations

As a part of my work for Global Action Plan’s EcoTeams project, I’ve been building reporting tools to predict household heating energy consumption into the future from some measured readings. This post is all about how to predict energy consumption based on a process of degree day adjustment.

(NB — what follows is a bit technical..)

The Carbon Trust succinctly say:

Degree-day figures quantify how hot or cold the weather has been as a single index number for the region and month (or week). They allow you to account properly for the effect of weather on energy consumption.

Projecting energy consumption for heating forwards involves some calculations — you need to consider changes in the outside temperature, and what impact that is going to have on the energy required for heating inside your building or home. When it is getting colder each month, the amount of heating and energy used for heating goes up. And when summer approaches (we hope) that the outside temperature goes up, and the requirement for heating drops away.

Each year, the weather is different, so the degree day values for each month or week change.

The meaning of degree day values

So, what do these degree-day numbers measure and how are they calculated? I’ll explain a bit.

There’s an assumption used here, that if the outside temperature is 15.5C, the building will be able to warm itself without needing to use energy for heating. Buildings are warmed by people, by heat from the sun, by the heat from equipment in the building among other things to bring the outside temperature up to a reasonable internal temperature.

A degree day is then calculated using the 15.5 degree value as follows:

degree day = 15.5 – outside_temperature * days

A weekly degree day value sets days above to 7, while for a month, it is set to the length of the month in days.

So, the degree day value is bigger when it is colder, and the degree day value is proportional to the energy required to heat the building to a normal comfortable temperature. This gives us the information we need to predict future energy consumption, or compare enery consumptions in different months even though the outside temperature was different.

An example

Say we wanted to work out our energy consumption for Nov 2007 compared with Oct 2007.

Let’s say in Oct 2007 we used 500 kWh heating the house. And in Nov 2008 we used 680 kWh heating the house. We were trying to reduce our energy consumption by turning down the boiler. Did we succeed?

So, we get the degree days values for South-East England:

Oct 2007: 166

Nov 2007: 248

Okay, we can immediately see that November was a lot colder than October, as you’d expect. So we’d expect our energy consumption to go up a lot. But let’s do the calculation:

energy_used_oct / dd_oct * dd_nov = predicted_use_nov

or

500 kWh / 168 * 248 = 738 kWh

So 738 kWh is our predicted energy use for heating adjusted for the relative warmth of the two months.

But we actually used only 680 kWh, so that means we’ve saved a fair bit by turning down the boiler.

Conclusion

So, using the degree days values we can make these calculations, and end up making much more reasonable comparisons between months than if we just take the raw kWh values. Very useful.

References

You can find some historical degree day data from the Carbon Trust (PDF doc) .