(second in a three-part series)
Mike Opitz PE, LEED AP
Certification Manager, LEED for Existing Buildings
U.S. Green Building Council
Recap
The first article of this series summarized what energy Measurement & Verification (M&V) is, and what it can do for building owners and facilities managers. In essence, it’s a standard set of ways to determine how much real savings results from an energy project, which is critical in determining a project’s true impact to both your bottom line and to the environment.
Determining true savings is a complex affair because energy savings can’t be measured directly: it must be derived by comparing energy use before and after the project. Doing that comparison properly can be a real challenge because of all the factors unrelated to your project that cause energy use to fluctuate over time. In the end, measured energy savings are meaningful only if the final value is larger than the noise (variations, or uncertainty) in the measured energy use before and after the energy project.
This second article will focus on the main issues involved in planning the M&V for your project—identifying the risks you face, having proper expectations about the reliability of your final savings number, and how your desired reliability will affect your M&V costs.
Energy Savings Risk—Knowing what you Face
Virtually all energy projects are undertaken with an expectation to save money on the utility bills (electric, fuel, steam, water). Part One discussed whether the bill actually goes up or down, and explained that for our purposes energy savings is best viewed not as a change in the bill but as avoided energy use or cost-a reduction in energy use compared to the what would have occurred without the energy project, assuming nothing else changes. But this begs the larger question: how can you be sure you’ll get real avoided energy use, and how can you maximize it?
All energy projects involve three main kinds of energy savings risk that limit the actual savings as well as your ability to know how large the savings is:
- Performance risk—whether the facility changes made in the project truly deliver the reduced energy use they are intended to deliver (e.g., does a new light reduce the power draw as much as expected? Does a new chiller operate as efficiently as expected?)
- Usage risk—whether the facility experiences the same conditions after the project as it did before the project (e.g., weather, operation hours, occupancy levels, heating and cooling setpoints)
- Uncertainty risk—whether the M&V process detects and quantifies the first two forms of risk, plus limitations inherent in the M&V process that lead to imperfect knowledge
Facility managers reduce performance risk by selecting and installing equipment carefully, and by using and maintaining it properly afterwards. Usage risk can sometimes be managed (HVAC setpoints), but is often uncontrollable (weather) or mission-driven (occupied hours, occupancy level), meaning the best you can do is monitor usage parameters so you’ll know their impact on savings. This is done with the savings adjustments discussed in Part One, which often require dedicated measurements on the usage parameters. Without these adjustments you won’t know the difference between the change in the bill and the true project savings.
That leaves uncertainty risk, which M&V is intended to reduce directly. The first step in dealing with uncertainty risk is deciding how much of it you can tolerate, which determines how much M&V you need and how much it will cost.
Planning the Amount of M&V—Certainty vs. Cost
M&V doesn’t come free, so your planning task is to meet the needs of the project within the available budget. The appropriate level of M&V effort scales with the size of the project’s energy savings; a typical range for M&V costs is 3-10% of project value. For a project of a given size, where your M&V costs land within that range depends on the complexity of the project’s savings measures, as well as how much you expect the usage parameters to change between the baseline period and the post-retrofit period. For a given size project, more rigorous M&V-measuring more systems, measuring a few systems in great detail, or developing more sophisticated savings models-will increase M&V costs. On the other hand, more M&V reduces the uncertainty in the savings, improving your knowledge of the results. How much M&V is enough?
There’s no single right answer here because any project’s needs may be unique. One approach is to assess your desire for certainty in savings and let that drive all other considerations. In the extreme case, if you need as much certainty as money can buy, then you’ll invest in cutting-edge instrumentation, fully measure every system affected by an energy project, develop a thorough engineering model of the energy use, and minimize the number of assumptions you have to make in that model. Such an approach would be very expensive, and for that reason it’s rarely done except for very small projects or if the measurements are valuable to the facilities staff for other reasons.
A second, more typical approach is limiting the uncertainty to a level you’re comfortable with, and investing in enough M&V to provide that level. The uncertainty expresses the relationship between the estimated savings, which is the result of the M&V process, and the actual savings, which will never be known with perfect accuracy. The more uncertainty you can tolerate, the cheaper your M&V will be.
Your uncertainty goal must be expressed in mathematical form to be meaningful, which requires that both of the following independent parameters be specified:
- Confidence level—how sure you want to be that the true savings value falls within the precision bounds you’ve specified (see next item); in other words, how repeatable the measurements are. Can be 70%, 80%, 90%, or any other value you specify between 0% and 100%. Higher is better.
- Precision level—the accuracy of the savings estimate, expressed as error bounds that you want the true savings to fall within. Can be +/- 5%, +/- 10%, or any other value you specify from 0% upwards. Lower is better.
These uncertainty goals are written in shorthand form as “confidence/precision”, e.g., 90/10, 80/20, so a hypothetical perfect measurement (impossible to achieve in practice) would be described as 100/0. The two uncertainty values are often selected such that they add up to 100%, but that’s not required: it’s valid to specify uncertainty of 90/20, 80/10, or even 72/45. What’s critical to understand is that both values are independent of each other and both must be defined: specifying either a confidence level or a precision level alone is mathematically meaningless, and means your target uncertainty is ambiguous. If that happens you won’t know if your M&V dollars are well-spent.
This topic is widely misunderstood and abused in practice, so you’re well advised to get comfortable with it. As one example, how many times have you seen equipment specs that claim to deliver performance within +/- 5% of some target value (precision), but without providing the associated confidence level? Whether the confidence is disclosed or even known by the spec writer, it does exist (i.e., it can be derived by knowing the testing procedure), and it is just as important as the precision. Achieving +/- 5% precision is unimpressive if your confidence in that range is only 40%!
If you choose this approach, how do you know what confidence and precision to specify? The answer is simple: whatever your project team is comfortable with. The more savings certainty you want, the more you’ll have to pay to get it. As a general guide, projects with uniform technology and predictable energy use (e.g., a lighting retrofit) can achieve high certainty at reasonable cost, perhaps 80/20 or even 90/10. Conversely, for mixed-technology projects or those with variable energy use (e.g., HVAC upgrades) the budget may only support 75/25 or 65/35. The stakes involved also affect this decision: you’d want to invest in more M&V if you’re paying contractors based on how much energy they save for you, but could do with less if the whole project is done in-house and its results won’t affect decisions about future projects.
A third, more formal approach to deciding how much M&V to do is to compare your project’s M&V “cost curve” (M&V cost rises as M&V quantity increases) directly with a similar curve showing the value of the energy savings that remains uncertain (which declines as M&V increases). You can strike a good balance by choosing the point where the two curves cross. At this point you are at or near the mathematically optimum M&V budget for your project because you have minimized your total “cost”.
Coming up in part III
Now that you know how to decide how much M&V you need, the final step is designing your M&V process: sorting through the standard technical M&V approaches and choosing the best match for your project’s traits, while meeting your uncertainty and cost goals. That’s what we’ll cover next time in the conclusion of this series.
Resources
International Performance Measurement and Verification Protocol (www.ipmvp.org)
Offers free downloads of several voluntary M&V standards representing consensus of global experts. Volume I includes an excellent introductory, non-technical overview of M&V concepts.
Federal Energy Management Program’s M&V Resources (www.eere.energy.gov/femp/financing/superespcs_mvresources.cfm)
Free download of the Department of Energy’s FEMP M&V Guidelines for federal government facility managers to use in planning and implementing M&V on energy projects, plus many other M&V planning resources.
American Society of Heating, Refrigeration, and Air Conditioning Engineers (www.ashrae.org)
Publishes ASHRAE Guideline 14: Measurement of Energy and Demand Savings. Highly technical, but a good resource for engineers responsible for executing or overseeing M&V work.
U.S. Green Building Council (www.usgbc.org)
The USGBC’s LEED for Existing Buildings program addresses M&V issues in several of its requirements and credits.
