You Say It Saves 20%. How Do We Know?

“We can reduce your HVAC energy consumption by 20%.”

Wonderful.

If you own or manage restaurants, retail stores or commercial facilities, you've probably heard some version of that before. The percentage might be 15%, 20% or 30%, usually accompanied by an impressive-looking graph.

And perhaps it's true.

But imagine you're the facilities manager who recommended the project. Six months later you're sitting across from the CFO and the obvious question comes:

“So, what did it actually save?”

“Should we roll this out across another 5 locations?”

Suddenly the percentage in the original proposal isn't much help. This is the problem Measurement & Verification, or M&V, is designed to solve.

HVAC measurement and verification - Measuring something that didn't happen

There's an odd thing about energy savings, you can't directly measure them.

We can measure how much electricity a building consumed. What we can't measure is how much electricity the same building would have consumed, at exactly the same time and under exactly the same conditions, if the energy efficiency project hadn't happened.

Fig: Estimating what energy would have been consumed in the absence of the efficiency improvements

We don't get to run February twice.

Once with the HVAC optimisation and once without it.

Instead, we have to determine what the building would reasonably have consumed without the change. This is sometimes described as the counterfactual.

At its simplest, the idea looks like this.

Simple enough on paper.

The interesting part is working out that first number.

A restaurant makes the problem obvious

Consider a quick-service restaurant (QSR) on a hot Sydney afternoon.

The kitchen is operating, refrigeration is running, customers are coming through the doors and the HVAC system is trying to keep the dining area comfortable while it's 37°C outside.

Now compare that with a mild Tuesday in May.

Same restaurant. Very different energy conditions.

Suppose we install an HVAC optimisation system and the next electricity bill falls by 18%. We'd obviously be pleased.

But how much of that reduction was caused by the optimisation?

Perhaps the weather was cooler. Perhaps trading hours changed. Perhaps customer traffic was different. Maybe refrigeration equipment was replaced during the same period.

Or perhaps the HVAC optimisation really did produce an excellent result.

Looking at two electricity bills doesn't necessarily tell us how much energy was actually saved due to our efforts.

And that's where M&V becomes useful.

Agree on the rules before the project starts

Rather than waiting until the end of a project and arguing about whose savings calculation is correct, M&V provides a framework for agreeing upfront on how savings will be determined.

The International Performance Measurement and Verification Protocol (IPMVP) provides an internationally recognised framework for doing this.

We agree on the baseline period and reporting period. We decide what energy data will be used and which conditions materially affect consumption. We document how changes in those conditions will be treated and how the final savings calculation will be performed.

Those decisions form part of a measurement and verification (M&V) Plan.

So when we eventually talk about a saving of 10%, 20% or 30%, both sides understand what that number represents, how it was determined and the methodology used to arrive at the result.

That's very different from deciding how to calculate the savings after seeing the result. Agreeing on the methodology upfront also helps avoid disputes later.

A baseline isn't just “last year's energy”

This is another important distinction.

Imagine our restaurant consumed 300,000 kWh last year. That's useful information, but by itself it doesn't tell us how the restaurant behaves.

A good baseline attempts to understand the relationship between energy consumption and the conditions that influence it. For an HVAC related project, especially a quick service restaurant, weather is an obvious factor.

Operating hours may matter too, along with other production variables that are relevant to the particular facility such as the floor space, fan airflow along other cooling and heating load factors.

So the question gradually changes from:

“How much energy did this restaurant use?”

to

“What drives this restaurant's energy consumption?”

Once we understand that relationship, we can make a fairer comparison during the reporting period, the period where we measure the actual energy usage with the efficiency measures in place.

If the reporting period happens to contain a much hotter summer, for example, we don't simply penalise the project because the HVAC system had more work to do and used more energy.

We account for the change in conditions according to the M&V methodology we agreed at the beginning.

So how do we calculate the savings?

For many of the HVAC optimisation projects we're developing, our starting methodology is IPMVP Option C , Whole Facility. This methodology places the measurement boundary around the facility.

Historical facility level energy data and relevant independent variables are used to establish a baseline relationship using statistical modelling.

The agreed baseline model is then used to calculate what energy consumption would have been under the conditions experienced during the reporting period.

The baseline data must be representative of a typical operating period, so if any "non-routine" events occurred during the measurement period, they are included as adjustments into our model.

A non-routine event could be a un scheduled change in production activity, significantly adding or reducing HVAC load through equipment changes or other material events that parties agree to that would otherwise distort the baseline measurement.

We are asking - What would this site reasonably have consumed under these conditions if the optimisation hadn't happened?

Then we compare that with what the site actually consumed. The difference is the avoided energy consumption determined under the agreed methodology or the "determined" savings.

From one restaurant to a whole portfolio

Now imagine the business operates 100 restaurants. We don't necessarily want to invent an entirely new approach 100 times.

QSR sites share characteristics. There are common operating patterns, equipment types, HVAC demands and business processes. That gives us the opportunity to develop a repeatable sector level M&V framework, a common starting point for understanding how this type of facility uses energy and how savings should be assessed.

But a framework isn't a substitute for understanding the individual site.

A restaurant in the eastern suburbs may trade longer hours. A restaurant at Penrith may have different HVAC equipment. Another site might experience very different cooling loads or have undergone a significant equipment change during the reporting period.

So we can start with a common template and then progressively make the analysis more specific to the actual facility.

• The framework provides consistency.

• The site provides the detail.

That becomes particularly useful when the eventual objective isn't simply proving one project worked, but understanding energy performance across an entire portfolio.

Fig : A conceptual view, based on internal systems, of how we calculate & report aggregated M&V based portfolio savings

The story doesn't stop on installation day

Once the optimisation is installed, the reporting period begins and the Option C methodology continues to do its job.

Actual whole-facility consumption is compared with the adjusted baseline under the conditions experienced during the reporting period. As more data accumulates, we begin to see how the result develops across different operating conditions.

This matters particularly for HVAC.

A few mild weeks after commissioning aren't necessarily representative of performance through a hot summer. An early result can be useful, but it shouldn't be given the same weight as a result supported by a broader and more representative range of conditions.

So we don't change methodologies as the project progresses.

The Option C framework remains the same.

The evidence supporting the savings determination grows over time.

At the same time, we're interested in more than the final savings number, we want to understand what's happening inside the HVAC system.

Post installation sub-metering and operational data can provide another layer of visibility into HVAC energy consumption, temperatures, equipment states, loads and control behaviour.

That gives us two complementary views of performance.

1. At the facility level, Option C helps us determine whether the expected reduction in energy consumption is appearing in the overall site result.

   
   

2. At the HVAC level, operational and sub-metered data helps us understand whether the system is operating as expected and whether that performance is persisting.

For example, if performance begins to change, the additional data may help us investigate why.

• Has HVAC load changed?

• Is equipment behaving differently?

• Have operating conditions changed?

• Or has something significant happened at the site that warrants investigation as a potential non-routine event?

The additional monitoring isn't a different savings methodology. Our agreed savings determination remains Option C, Whole Facility. Instead, it gives us a better view of the physical system behind the result.

Put simply

Option C helps us determine the savings.

Ongoing monitoring helps us understand and maintain them.

That becomes increasingly valuable as one site becomes ten, fifty or a hundred.

Where this gets interesting

Historically, M&V could be quite labour intensive. Data was collected, models were built and reports were produced periodically.

Increasingly, we have access to interval meter data, HVAC sub-metering, sensors, cloud platforms and analytics that can process information much more frequently.

This is where ideas such as Advanced M&V and M&V 2.0 start becoming relevant.

The fundamentals don't disappear. We still need a credible baseline. We still need to understand changing conditions. We still need an agreed methodology and appropriate data.

What changes is our ability to apply those principles more frequently and potentially across much larger portfolios.

Imagine moving from:

“Did this restaurant save energy last year?”

to

“How are all 100 restaurants performing right now, and where should we look next?”

That's a much more interesting question.

And it's where we think M&V can evolve from being something used primarily to verify a completed energy project into something that helps businesses continuously understand and manage energy performance.

So, about that 20%...

Let's return to where we started.

“We can reduce your energy consumption by 20%.”

Perhaps.

But before the project begins, there are some more useful questions to ask.

• What's our baseline?

• What affects energy consumption at this site?

• What are we going to measure?

• How will we account for changing conditions?

• And how, exactly, are we going to determine the savings?

If the client and supplier can agree on those things before the project starts, we've already solved one of the biggest problems in energy efficiency.

We're no longer asking the client to simply believe the claim.

We've agreed on how we're going to test it.

More on measuring energy performance

This article is part of an ongoing look at practical Measurement & Verification for commercial HVAC and multi-site businesses.