Are You Really Using Less Electricity? Just Look at the Weather

Share Button

An environmentally conscious energy customer plans to install LED light bulbs throughout his family’s home. It’s an investment, he knows—these bulbs aren’t cheap. But he figures it will be worth it because he’ll surely reduce his energy use.

Or will he?

It seems like a simple question. A quick comparison of pre- and post-LED energy bills would appear to give an answer, demonstrated by dollars saved from one billing cycle to another. But the comparison won’t give a complete picture of changes in the customer’s energy consumption patterns unless it also accounts for one very important factor: the weather.

Here’s how weather makes a difference. The customer receives a bill for $50 for this July’s post-LED installation energy use. He compares it to last July’s pre-LED $70 statement, and he can easily attribute this savings to the new bulbs. But what if the average temperature this July was a mild 75 degrees, and last July it was a sweltering 90 degrees? How does he know he didn’t really save that money by using less air conditioning this summer? If this year’s average had been 90 degrees again, what would his savings have been? How much difference would the LEDs really have made?

A simple question has suddenly become quite complex.

Weather normalization leverages real-time data

Luckily, energy providers can help their customers get to the bottom of the question: “am I using less electricity?” by incorporating weather normalization into their customer reports.

Weather normalization uses statistical science to correlate energy consumption to up-to-the-minute weather conditions. In doing so, energy providers can determine what the customer’s bill would have been if weather didn’t vary month-to-month and year-to-year.  By removing the impact of weather, the energy provider can now provide a far more accurate view into how much energy the customer saved compared to what he would have spent had he not made any changes.

Real-time analysis means better customer personalization

By factoring in the unique weather patterns around their homes, customers can receive individualized results that tell them exactly how they’re doing in terms of managing their energy consumption and the associated costs. Instead of receiving incomplete comparisons to last year’s consumption, energy providers can offer their customers tailored information that instead tells them, for example, “You were six percent more efficient this year, which means you saved $11 on your bill.” Those numbers will be accurate no matter how different temperatures were from one year to the next.

In addition to giving customers this meaningful assessment, weather normalization can help energy providers target offers for additional products and services to individual customers. When these energy saving tools are tailored to customers, presented in the ways they prefer to engage and made relevant to their specific environments, customers are more likely to participate, benefit financially and reduce their energy use.

And for the customer who installed LEDs in an effort to use less energy? He can assess more precisely the impact of that investment. He can know that the energy use reports he receives post-installation don’t exist in a vacuum: they reflect energy use within the existing conditions and they compare it to what energy use would have been in the same weather, pre-LED adoption. The customer will know for sure whether the investment was worth it, and he’ll know exactly what to do next to keep reducing his energy consumption.

Chris Black is CTO/COO of Tendril.


Share Button


  1. Specifically for the example of LEDs, the much bigger factor in answering the question “did they work?” is how long the lamps were on, possibly affected by number of people in the household from year to year (lights on in multiple rooms), and whether any of these individuals changed the amount of time that they were home during the two time periods.

    Even if the comparison to temperature may indicate that there wasn’t a savings after accounting for different temperatures during the two time periods, it wouldn’t mean that the LEDs are defective — they might have been run longer, or there could be some offsetting effect from elsewhere in the house from an electric appliance.

    The real-time data could be useful in that regard, to show a daily power profile that should reflect occupancy when lights or other appliances are used, and identify any periods when power use is high but the house is not occupied.

    Weather normalization will be a much more powerful tool in evaluating the effectiveness of heating and cooling system modifications, low-flow showerheads (by deducting the weather-related portion of energy use and comparing what is left between the two periods to show improvement), or weatherization measures.

    Savings from LEDs might be visible from weather normalization of the electricity use, particularly if the location of the house has a time of year when neither heating or cooling is used. Then the difference in energy use can more reliably be assigned to lights or other appliances that run all year.

    • Les Lambert says:

      The question Chris poses comes up for all sorts of retrofits, as well as for operating and maintenance measures intended to save energy. The basic question is, “Did it work?” and can I tell from the utility bills alone? I’ve developed a method I call “diagnostic benchmarking” to answer such questions (see

      David Eldridge poses another key question that is pertinent, in addition to weather. My method works for situations where the building or facility use is consistent from one week to the next. In the commercial building context, having a building automation system promotes consistency from one week to the next. In a residential context, using a programmable thermostat and “groundhog day” habit patterns will do the trick.

      A third question is interactive effects between whatever is done to reduce energy use, and energy used for HVAC. In Chris’ example, the answer could be that LED lighting also reduces cooling energy, in a cooling-dominated climate – a win-win outcome. Alternately, in a heating-dominated climate, where the heating means is electric forced air heat (which I don’t recommend), the LEDs will probably have negative energy savings.

      I wish more utilities would help their customers with such issues. They have (or can obtain) all the data needed to rigorously address such questions.

      Les Lambert P.E.

  2. If a good base line is established before an energy project, one should be able to see the effects of any Energy Conservation Measure (ECM). An straightforward way to establish a project baseline is to establish a “daily baseline” for each month of the year. This involves averaging three years of utility history, assuming there is not a trend up or down over that three years, then dividing by the number of days in each months billing cycle. This takes out one of the biggest variables, number of days in each billing period. The variables left will be weather and occupancy.