Today, let’s consider a subject that is as squishy, subjective, and amorphous as net savings and non-energy benefits (see here and here). Today’s subject is savings persistence.
Savings Persistence and Its Importance
Even allowing for generous latitude, if I polled readers of this post, I would probably get a dozen definitions of savings persistence. For retro-commissioning, persistence is ensuring the measures aren’t undone. For the broader group of behavior programs, persistence is getting customers and their occupants (family or employees) to continue to value and manage energy over the long term.
To the evaluator, savings persistence opens many cans of worms. The definition boils down to what are the savings over the life of the measure compared to the baseline alternative. The lifetime (years) of the measure is even part of the equation itself.
Determining persistence is critical for benefit cost tests. The benefits of energy efficiency are the present value of all future savings and non-energy benefits.
A good source of persistence information is included in the Uniform Methods Project (UMP), but I will condense, exemplify, and expand on some of that.
Changes in Operating Hours
Assume a manufacturing facility operates three shifts 24/7. Demand for this customer’s product softens, and they cut back to partial two-shift operation. Do the savings increase or decrease? It would seem to have obvious answers, but no.
Consider a lighting project. The lighting hours of use should decrease, but do they? If manually controlled, are they shut off? Likely to some degree, but not completely. What about motion sensors controlling the lights? Now they are definitely being shut down more. Do savings increase because the controls produce more benefit? The real answer, not to be confused with the official answer, is yes.
What about a new air compressor in such a situation? The new air compressor has much more favorable part-load performance compared to the status quo. Does the compressor save more energy as a result of the cutback? The real answer, not to be confused with the official answer, is yes.
A 7:30 pace for a marathon is as difficult as a 6:30 pace was for me twenty years ago. Am I less fit? Gee willikers, according to the Boston Athletic Association, the answer is no. The qualifying time is almost exactly a minute per mile slower due to the age difference. However, one could argue that I’m more fit because the status quo has inched closer to cardiac arrest, stroke or type two diabetes, than I have, all else equal. So it is with equipment.
Some equipment is more efficient at beginning of life. Cooling equipment is one good example. Over time, fins on air-cooled condensers get dirty and damaged. Tolerances on compressors grow making them less efficient. But the same things would happen with the less efficient alternative in many cases. While the efficient equipment loses efficiency over time, the gap with the alternative may actually grow, saving more, in theory, over time. The more you spend, the more you save.
Savings persistence is a function of equipment life and durability, but do we even want equipment to last for its expected life? According to New York and California technical reference manuals, and ASHRAE, the condensing unit for my AC system died last year. It is very much alive as it kept us cool through the past couple nights of sauna-like weather conditions.
Is it a good thing that this now-outdated and inefficient unit is still running? More perversely, if I did replace it, conventional evaluation would declare it was beyond its useful life, and therefore the minimum baseline is actually 13 SEER rather than the 10 SEER as is. It could last another 20 years, and when I say “could”, I mean it will unless I decide to replace it before it dies.
Performance Degradation (and Deficiency)
Performance degradation includes classical control and operational factors. Performance deficiency is my primary gripe with new construction programs and code compliance issues. The stuff you can touch and see is there, but the building performs woefully because of what you cannot see without the necessary deeper look – control sequences and design factors.
Again, I refer to the more you spend the more you save. That is, the more poorly certain equipment is controlled, the more it saves relative to a less efficient alternative, maybe. For example, if lighting controls don’t work they are not saving energy; but the efficient lighting being controlled (or not controlled) is operating longer and thus saving more than initially estimated. Per the UMP, if manufacturing cuts from two shifts to one and hours of use decrease by half, savings are reduced by half. I disagree. Otherwise, we must reward bad behavior with more savings as I just demonstrated.
Persistence factors can be categorized in two clumps: hard and soft.
- Hard persistence factors include clean or measurable change: equipment degradation, changes in operating hours, and equipment life.
- Soft persistence factors include degradation due to behavior: measure is undone shortly after implementation or slowly undone for something like a facility-wide energy management system or home energy reports.
Just as attribution studies include free riders and spillover, persistence includes degradation and advancement. Savings can increase due to differences in degradation of alternatives over time, or behavior can advance savings with continuous improvement over time.
In either case, persistence studies should consider how supply side assets would be factored.