Measuring at the Meter: California Takes the First Steps

california logoEnergySavvy’s Jake Oster explains how California’s shifting rules for measuring at the meter open the way remake energy efficiency and pay-for-performance programs.

A few weeks ago the Governor of California signed AB802 into law. AB802 is a first-of-its-kind energy efficiency law that calls for the measurement of energy savings to be evaluated based on “normalized metered energy consumption.” As I wrote when the bill passed, this is a major milestone for the measurement of energy efficiency because measuring at the meter, while making adjustments to account for outliers, is a true accounting of how energy efficiency impacts the grid and provides value for ratepayers.

The first step in implementing AB802 is coming quickly, in the form of “high opportunity projects and programs,” which go live in early 2016. To set the table for a 2016 roll out, last week the California Public Utilities Commission (CPUC) released its first indication of what the new rules for measuring at the meter will look like.

“Normalized Metered Energy Consumption”


At the core of the new law are these four words and how they will be interpreted. For those of us in the industry, these four words can mean many different things, but it is the CPUC’s interpretation that matters the most. And the first draft of the CPUC guidance is a promising start.

First, let’s take the most open-ended term, “normalized,” and what this means. According to the CPUC, we can’t ignore the other factors that can impact energy consumption. Normalized energy consumption has to account for the external factors that “can influence energy use trends.” In practical terms, that means variations in rates, commodity prices, or even macroeconomic changes across a service territory. For example, if natural gas prices suddenly rise, a reduction in natural gas consumption is a likely result. You don’t want to accidentally count those reductions as energy efficiency. Accurate measurement should control for these types of external factors and is an important part of providing an accounting for energy efficiency that will be accepted by all stakeholders.

The other three words are a little more straightforward but still require some important parameters. “Metered” makes clear that savings must be tied to a meter. For example, point-of-sale retail programs that allow customers to buy discounted lightbulbs at the hardware store are not eligible for metered measurement. Because those savings cannot be tied back to a meter, we will have to continue measuring those programs with predetermined estimates. Savings also needs real meter data. While estimated meter reads may be acceptable for billing, real meter data is a requirement for measuring energy savings. Real meter data is an important factor for getting measurement correct. Finally, the simplest phrase, “energy consumption” accounts for a change in energy usage as the result of an intervention.

For measurement policy wonks, the important signal is that the CPUC’s approach carefully balances rigorous measurement protocols with the practicality of quickly rolling out high opportunity projects and programs. The starting definition for “normalized metered energy consumption” protects ratepayer dollars and allows California utilities to move quickly and nimbly to roll out programs under the new measurement paradigm.  The CPUC is also ensuring that transparency remains a centerpiece of energy savings measurement. Black box approaches won’t meet the test. Considering the importance of getting metered measurement “right,” measurement methodologies will need to be replicable, open to peer review, and well documented. This level of transparency will be critically valuable to an industry that is learning and innovating in this new paradigm for valuing energy savings.

Pay for Performance

In the same document that defines metered measurement, the CPUC has also taken the first steps to outline how pay for performance programs can move forward. Tying together pay for performance and the measurement of metered savings makes good policy sense. Therefore, it is important the rules are written in tandem.

To incentivize energy savings, utilities, contractors, and implementers will need to measure energy savings at the meter and use “normalized metered energy consumption” as a framework for proper energy efficiency measurement. In order to put a value on energy savings, metered measurement must follow clear protocols that protect ratepayers, program participants and industry stakeholders who implement energy efficiency programs. Ensuring measurement is “normalized” and controls for external factors is the type of policy expertise that will provide accurate and defensible measurement to support pay for performance programs.

The Start of a Process

This document is the first step in a long process to truly define metered measurement and pay for performance programs. The CPUC will be hearing from stakeholders on this guidance and this is only the initial phase for high opportunity projects. As California takes the first steps toward metered measurement for energy efficiency, there will be lots of voices at the table and changes shaped by those differing perspectives.

At EnergySavvy, we plan to be part of the discussion. While getting metered measurement perfect is probably impossible, getting it right is required for this experiment in measurement to succeed. The value of energy efficiency is built on the bedrock of measurement, and the future of energy efficiency is built on programs that value energy savings as a resource. Measuring savings at the meter ties those strings together and enables energy efficiency to serve as a resource for the industry and the climate. And one thing is certain: if California is remaking energy efficiency measurement, it is going to impact the entire industry and reverberate around the country.

Jake Oster is  EnergySavvy’s senior director of regulatory affairs. This blog originally appeared on EnergySavvy’s website.

Sign up for our newsletter and get the latest microgrid news and analysis.
Subscription Type


  1. I strongly disagree with the authors interpretation of normalized metered savings as including controls for societal changes. This kind of thinking will put us right back in the uncertainty that is inherent in current EM&V methods, and will prevent efficiency from being a reliable resource that can attract private investment, and specifically I do not believe this was the intent of the CA legislator.

    For METERED efficiency to work in the market, the way we calculated savings needs to be like an actual meter, meaning it can be replicated by all parties in a transaction. For this to be true, you simply can’t include control group studies, which cannot be predicted in advance or replicated (market actors can’t create the same controls that a consultant or regulator might come up with in the future) therefore this kind of thinking casts uncertainty on the measurement of efficiency.

    A Meter is defined as “an instrument for measuring, especially one that automatically measures and records the quantity of something, such as gas, water, miles, or time,” (

    You can’t automatically measure efficiency if you make these sorts of adjustments after the fact. If you do, then private investment in efficiency as a resource will continue to be uncertain and unpredictable – and both those attributes are abhorrent to private investment.

    What is being proposed in this article is in fact just an automation of current flawed EM&V methods, and the goal and intent of AB802 and SB350 is a much more fundamental change.

    These sorts of non-routine adjustments for external factors, in a model where efficiency is treated as a procurable resource not just another program, should be handled not in how we count saved electrons or BTUs, but instead in how they are valued. If savings would have happened anyway, then there will be more supply, and in markets where supply and demand function, the price will drop.

    Grid planners acquiring efficiency will also take these factors into account in how much savings they acquire, which will also impact the value. If one determines that 20% of savings were free riders, that does not change the reduction from baseline on each building, it instead means that 20% of the savings are already in the utilities baseline, and therefore, in order to keep the lights on, they must acquire 20% more savings to achieve the same load shape.

    I believe that this interpretation of Normalized Metered savings sells us all short. The goal of AB802 and SB350 is not an automation of how things work today, but an attempt to break free of current regulations and inside the box regulatory thinking.

    Step one is that we start to count our savings in a way that is transparent and replicable so that we can all agree and have confidence in the product of our work. Step two is determine the value of efficiency as a resource realtive to the cost of other distributed resources such as storage, DG, DR, etc. Control groups have a place in establishing what efficiency is worth, but simply cannot be part of how we “Meter” results… the private sector needs certainty not more subjective EM&V reports that change numbers after the fact.

    For a view into what the California legislator specifically had in mind I would suggest checking out the NRDC and TURN proposal submitted to the CPUC last spring.

    These bills are demanding a fundamental change in how we account for energy efficiency, and how we go to market with efficiency. We need to step outside the current regulatory box and change with the times. The future of efficiency is moving beyond programs towards markets, and the first step to enable that is to agree on how we measure or product with certainty. This approch being proposed by this author simply does not get us there.

Leave a Comment