Why Cutting Insulation Backfired on a 57-Unit Worcester Project

A 57-unit project in Worcester, Massachusetts thinned the insulation to save money. Walls dropped to R-12.6 continuous exterior insulation with no cavity batt, and the roof went from R-60 to R-50. Both changes removed real material and real labor, and both showed up as savings on the estimate. To still meet code, the team swapped electric tanks for heat pump water heaters. Those heat pump water heaters cost more than the insulation saved.

The Real Envelope Cut

Start with the wall, because it is the larger of the two moves. R-12.6 continuous exterior insulation with no cavity batt means the entire thermal performance of that wall sits in one outboard layer. The stud bays are empty. Whatever the wall does, it does with the continuous layer alone.

That is not a small edit to an assembly. It removes an insulation trade from the building, along with the batt material, the installation labor, and the inspection that goes with it. A wall with nothing in the cavity is also a simpler wall to build and a faster one to close up, so the savings are not only material.

The roof cut looks bigger than it is. Moving from R-60 to R-50 takes ten points of R-value off an assembly that was already deep. In heat flow terms that shifts the roof from roughly 0.017 to 0.020 in U-value, which is about twenty percent more heat moving through the roof, applied to the smallest exterior surface on a multi-story building. It is a real loss. It is not the loss that moved the model.

The wall is where the energy went. Continuous exterior insulation with an empty cavity is a clean detail and it handles thermal bridging well, but at R-12.6 there is not much of it. On a building with this much wall area, that is the change the model felt.

The Compliance Backfill Problem

Massachusetts multifamily compliance runs on a performance budget. The building has to hit a target, and the design team gets to choose which measures pay for it. Envelope, windows, HVAC, ventilation, and domestic hot water all draw on the same account.

Cutting insulation spends that account. It does not reduce the target. So the moment the walls and roof came down, the project had a gap, and something else on the list had to close it.

The team went to domestic hot water, and that choice makes sense on its own terms. An electric resistance tank is a one to one appliance, putting out about as much heat as the electricity going into it. A heat pump water heater moves heat instead of making it, so it delivers the same hot water for a fraction of the energy. In a well insulated multifamily building, hot water is one of the largest remaining loads, which makes it one of the largest available levers.

That is why the swap worked. It closed the gap the insulation cut opened. The question was never whether the backfill would reach compliance. It was what the backfill cost.

Why the Backfill Cost More Than the Cut

The two sides of this trade scale on different units, and that is the whole story.

Envelope savings scale with enclosure area. The building has a fixed amount of wall and roof, and thinning the insulation saves some amount per square foot of it, one time. Add apartments to the building without changing the footprint and the envelope savings barely move.

Domestic hot water scales with unit count. Fifty seven apartments need fifty seven water heaters, and every one of them absorbs the price difference between an electric resistance tank and a heat pump water heater. The envelope cut was paid once across the enclosure. The equipment upgrade was paid fifty seven times.

The equipment premium is also more than the appliance. A heat pump water heater is physically larger than the tank it replaces, it needs clearance and air volume to work, and it produces condensate that has to go somewhere. Depending on where it lands in the unit, it can pull in louvers, ducting, a larger closet, a condensate line, and coordination with the mechanical and plumbing scopes that an electric resistance tank never required. Some of that is drawing work. Some of it is square footage. All of it arrives per unit.

So the project traded a one time saving spread across the enclosure for a recurring cost multiplied by the number of apartments. On a 57-unit building that arithmetic does not come out ahead, and it did not here. The insulation cut saved money. The compliance it broke cost more to buy back.

The second order effect is worth naming too. The building that came out of this exercise has a thinner enclosure and better equipment. That is a worse trade than it sounds, because insulation keeps working with nothing plugged in and no maintenance, while equipment has a service life, a filter, and a replacement date. The team paid more for a building that models the same and holds up less well over time.

When Thinning the Envelope Still Pays

When the model shows real margin. If the design is already clearing its target with room to spare, an envelope trim can be free. The cut only costs money when it forces a backfill, and a project sitting comfortably above the line does not need one. The only way to know which situation you are in is to have modeled the building before the value engineering meeting rather than after.

When the roof is the only layer being trimmed. Roof insulation is deep into diminishing returns on a code minimum multifamily building, and roof area is small relative to wall area once you are above three or four stories. Trimming R-60 to R-50 on its own is often absorbable. It was the wall that did the damage here.

When the enclosure is large relative to the unit count. The arithmetic above inverts on the right building. A low rise with a big footprint and few apartments has a lot of enclosure per unit, so envelope savings are large and per unit equipment penalties are small. Townhouses and small garden style buildings behave very differently from a 57-unit stacked flat building.

When the backfill is already in the design. If the project already specified heat pump water heaters, the domestic hot water lever is spent, and an envelope cut has to be paid for somewhere more expensive. That sounds like an argument against cutting, and often it is, but it also means the comparison is honest from the start. Nobody discovers the real price of the cut three weeks later.

When the cut removes a whole step instead of a layer. Going from a filled cavity to an empty one takes a trade off the wall, and that carries schedule value beyond the material. A trim that only makes an existing layer thinner saves material and nothing else, which is a much weaker case.

The Decision Framework

Price the cut and its backfill as one line. The insulation reduction and the water heater upgrade were the same decision made weeks apart, and they only look like savings if you stop reading after the first one. Any envelope change that touches compliance has a second half, and the second half belongs on the same page as the first.

Run the model before the value engineering meeting, not after. By the time the walls are drawn at R-12.6 and the roof is at R-50, the compliance gap is a fact and the team is shopping for the least bad way to close it. Modeling the change in advance turns that into a choice between two priced options instead of a scramble.

And take Worcester as this building's answer rather than a rule about insulation. Change the unit count, the ratio of enclosure to apartments, or how much margin the design started with, and the same cut lands somewhere else entirely. There is no envelope spec that is correct across projects, because there is no project that is much like the last one. The number that matters is the one your building produces, and you only get it by modeling your building.