At year’s end, it is natural to reflect on the past 12 months, adjust course and plan next year’s goals. In the energy management world, measurement and verification (M&V) is a critical tool to determine if savings—or avoided costs—were realized. From these measurements, you can determine if you are on track or need to plot a new course.
This article will explore how M&V is evolving and new opportunities for its use, partly due to big data.
The Origin of M&V
More than 20 years ago, the International Performance Measurement and Verification Protocol (IPMVP) was created to establish fair ways to measure and verify savings. This is especially useful in buildings where all kinds of changes (occupancy, area, weather, etc.) occur over the years. The IPMVP guidelines provide detailed examples and step-by-step instructions/calculations. Information on IPMVP is available here: https://evo-world.org/en/
The IPMVP isa framework for developing an M&V plan for a specific project. Followed properly by all parties involved, a well-written M&V plan should result in a successful project with mutually agreed-upon results that are as fair as possible to all. But the process is still cumbersome and is usually applied only on larger projects.
In recent years, technical innovations—including advanced metering infrastructure and new data science—have created new opportunities to expand the application of M&V and bring down the cost of implementation, However, even free and ubiquitous metered data solves only half the problem. M&V requires that metered energy be compared to a modeled baseline—and not all buildings and their energy loads behave in consistent ways in the short or the long term, complicating the creation of satisfactory models.
Despite the data sources and analytics available, there is not yet a simple, low-cost solution that addresses the fact that buildings don’t behave consistently. The current M&V process requires a qualified individual to assess the details of each situation and create a custom M&V plan that attempts to balance cost and accuracy.
Even the best plans need to be updated over time because tenants leave, data centers are installed, growing trees block the sun, etc. Any number of changes can complicate the M&V process, requiring agreed-upon methods for making non-routine adjustments to the savings calculations. Most of these adjustments typically require some data collection and expert analysis, adding costs that can swamp smaller projects.
Simply explained, M&V 2.0 is about using the more abundant data in smarter ways. It is big data applied to savings calculations. A recent best seller in the area of predictive analytics, The Signal and the Noise, reviews how different industries are adopting two types of forecasting tools to their businesses:
- Statistical models based on measurements of energy data and variables that influence energy use. ASHRAE calls these inverse models.
- Physical models based on the physical characteristics of energy consuming-equipment and loads and variables that influence energy use. ASHRAE calls these forward models.
There has been great progress in both modeling techniques. M&V practitioners can access free, open-source tools that are much more powerful than what was available when the IPMVP was first conceived.
Obstacles to Hurdle
It is enticing to envision a world where low cost analytics from both statistical and physical modeling will be ubiquitous, but there are barriers.
- Perhaps foremost among them is the fact that having access to more sophisticated methods does not guarantee consistent results. Instead, the more sophisticated the modeling, the more likely that two practitioners will arrive at different results. Although any two sets of metering results would agree within a very narrow margin (<1%), models generated by two sophisticated algorithms may vary by much more (>10%), depending on the assumptions.
- Additionally, if the software used is not open source and the methods transparent, reasonable people can voice doubts about the whole process. Any erosion of trust in the quantification of results leads to reduced investment in the industry.
- We still have the issue that buildings change over time, requiring non-routine adjustments. These calculations are currently custom engineered. Standard methods for identifying and quantifying non-routine adjustments will be required to scale the industry.
IPMVP has become the global standard, making communication of core M&V concepts more accessible to anyone wishing to quantify energy efficiency projects.
Although automated M&V promises some gains, it will not answer all of the situations. A synergy of new methods and a workforce trained in their use will need to be created. Recent conferences have shown a large number of strong papers demonstrating the use of M&V in all kinds of challenging situations. These conference session rooms are typically full of attendees asking good questions about how to use the current protocols to their best advantage. Investors—who demand M&V prior to investing—are beginning to ask the right questions, and understand the arcane language used in M&V circles.
Fundamental research and pilot programs are breaking new ground and revealing new opportunities on which to build the next generation of M&V. However, much work needs to be done to adapt these tools to meet the challenge of the new energy infrastructure and scalability. But the opportunity is there and the rewards are great!
Most of this article was contributed by Steve Kromer, an M&V expert who has taught M&V all over the world. Steve can be reached at [email protected] or 510-847-8535.