Foam board insulation beats fiberglass in most commercial assemblies I've audited—but only when every seam is sealed with a silicone joint sealant rated for the movement. Skip that detail, and you lose more performance through the gaps than either insulation option can recover.
I'm a quality compliance manager. I review material installations—roughly 200+ unique items a year—and I keep a front-row seat to the same failure: the insulation was fine, but the sealant choice was an afterthought. The most frustrating part? It's entirely avoidable. You'd think a written spec would eliminate the problem, but interpretation varies wildly from one crew to the next.
So the bottom line, before anything else: insulation choice matters less than sealing strategy. Foam board is the higher-performing option when you can detail the seams properly. Fiberglass is more forgiving on cost and moisture handling—but it leaks air at the perimeter if the sealant work is sloppy. Whichever you choose, the silicone sealant determines whether your envelope actually performs.
Foam Board vs. Fiberglass: The Real Difference
Let's start with what each material actually does, because the two get conflated far more than they should.
Foam board (polyiso, XPS, or EPS) is a rigid panel with a closed-cell structure. It delivers continuous insulation and a high R-value per inch. It's also a vapor retarder by nature. That's both an asset and a risk: it stops air movement, but it can trap moisture if the wall assembly doesn't have a drying plan.
Fiberglass batt is porous and semi-rigid. It slows heat transfer using trapped air, but it doesn't stop airflow. And the 'fiberglass is a barrier' thinking comes from an era when building science was simpler. Today we know better: fiberglass slows heat transfer. It does not stop air.
The practical consequence: a fiberglass wall with unsealed edges performs measurably below its rated R-value. And an unsealed foam board wall leaks air around every seam. Neither option is 'better' without a sealing strategy.
Why Silicone Is the Sealant Family I Trust
Here's what the insulation debate misses: performance depends on airtightness. And airtightness depends on the sealant at joints, transitions, and penetrations.
I've specified all the major sealant chemistries over the years. Polyurethane is strong but unforgiving to apply. Acrylic is cheap and paintable but doesn't handle movement well. Butyl rubber stays flexible but breaks down under UV. For exterior joints and insulation interfaces, silicone is the only family I trust for the long term.
The reason is the silicone polymer itself. The backbone is alternating silicon and oxygen atoms (Si-O)—not the carbon-carbon chain of organic polymers. That Si-O bond is inherently more stable.
The Si-O backbone resists UV degradation, stays flexible at -60°C, and maintains structural integrity at 200°C. No organic sealant polymer makes all three claims.
This is also where I need to kill a myth, once and for all. The 'all silicones are the same' idea was semi-true thirty years ago, when the market had fewer grades and most products were commodity items. That's changed. Formulations today are wildly different: one-part vs. two-part, addition-cure vs. condensation-cure, acetoxy vs. alkoxy. Calling something 'silicone' is the start of the spec, not the end.
Three Dow Corning Products I Spec Over and Over
I routinely write three Dow Corning part numbers into my specifications. The mapping is simple.
Dow Corning 3145 RTV Silicone Adhesive — for Bonding
If you're attaching foam board to a substrate or bonding any component that will face stress, 3145 RTV is the product. It's a one-part, room-temperature-vulcanizing adhesive. No mixing. No two-part inventory. Open, apply, assemble.
The numbers matter: adhesion to glass, aluminum, steel, thermoplastics, and most cured silicones. Service temperature from -60°C to 200°C. A UL 94 V-1 flame rating. And it's listed against MIL-A-46146, which is a handy consistency benchmark for anyone who has read too many vague data sheets.
As an inspector, predictability is what I value. When a contractor submits 'Dow Corning 3145 RTV' as the adhesive, I know exactly what to check—the data sheet gives me published properties for tensile strength and elongation, and batch-to-batch consistency is exceptional. I've never rejected a batch of 3145 for off-spec cured properties. I can't say that about any generic 'silicone adhesive' on the market.
Dow Corning 888 Silicone Joint Sealant — for Movement Joints
For construction joints that will move—exterior facades, the perimeter of insulation panels, transitions between dissimilar materials—I spec Dow Corning 888. It's a one-part, low-modulus silicone joint sealant.
Low modulus means it stretches instead of fighting the joint. A stiff, high-modulus sealant behaves like a rigid strap across a gap: when the joint moves, one end lets go. The 888 gives first, so the bond stays intact. It's designed for ±25% movement capability, meeting ASTM C920, Class 25. That standard-and-class combination tells me more than any brochure.
The adhesion list is broad too: concrete, aluminum, galvanized steel, glass, wood, and most coated surfaces. That makes it the natural choice for sealing the gap between foam board and structural framing, or between fiberglass-lined ductwork and a wall penetration.
Spray Silicone — Maintenance, Not Construction
Spray silicone is a different animal, and let me be blunt about it: stop trying to use it as a sealant.
It's a light lubricant suspended in a solvent carrier. It stops squeaks, conditions gaskets and weather seals, protects rubber from drying, and serves as a release agent on molds. It belongs in the maintenance kit—not in the construction specification.
I've seen this mistake twice in the last year: someone attempts to fill a small gap with spray silicone, expecting it to behave like a sealant. It doesn't. The solvent evaporates and leaves a thin oily film. It might quiet a squeak for a while. It won't stop water or prevent a draft. Use spray silicone for what it is—a convenience lubricant with a short service life.
How to Decide: Foam Board or Fiberglass?
Here's the decision framework I apply on actual projects:
Choose foam board when:
- You need continuous insulation without thermal bridges.
- You can guarantee proper seam detailing—backer rod, tooled sealant, and a crew that understands why it matters.
- Your moisture plan doesn't depend on the insulation drying out.
Choose fiberglass when:
- Material cost is the primary constraint and R-value per inch isn't critical.
- The assembly needs to breathe or dry to one side.
- You're filling irregular cavities where rigid panels would create gaps.
One pushback to the 'foam board is always better' crowd: it isn't, when the seams are treated as an afterthought. In our Q1 2024 quality audit, we caught a cold-storage addition where the installers had placed foam board with zero seam sealing. Interior moisture moved through the gaps and condensed inside the wall. The vapor-retarding nature of the foam trapped the water against the exterior sheathing. The remediation ran the contractor about $22,000—replacing panels, drying the cavity, re-insulating. The panels weren't the problem. The sealing was.
Even after we specified foam board for that correction, I kept second-guessing. What if fiberglass would have been more forgiving? The weeks between the audit and the completed remediation were stressful. In hindsight, that project gave me the rule I now use everywhere: any insulation will fail at unsealed joints.
When Silicone Is Not the Answer
An honest inspector states the boundaries. Silicone isn't universal.
- Deep joints. If the joint is deeper than about 12 mm (1/2 inch), install backer rod first. Silicones need controlled depth; a deep, unbonded joint will trap cure byproducts and form bubbles.
- Paintability. Most silicones won't take paint, and 888 is not among the paintable exceptions. If the finished surface needs to match, plan for that.
- Continuous immersion. Weather resistance under rain is not the same as submerged service. Check the immersion statement on the data sheet.
- Plasticized substrates. Some plastics and composites migrate plasticizers into the bond line, which can break adhesion over time. Verify compatibility first.
And one more: cheap 'economy' silicone is a trap. It looks identical in the tube, but it often has lower solids content and no published movement rating. It shrinks, it splits, and it fails. That's not an argument for always buying the most expensive option—it's an argument for checking the data sheet and the standard it claims to meet.
Five Questions Before You Approve a Sealant
I'd rather spend ten minutes helping a customer understand a specification than deal with a mismatched project three months later. An informed customer asks better questions and makes faster decisions.
Before you approve any silicone sealant or adhesive, ask these:
- Which standard does it meet? (ASTM C920, ISO 11600, UL, MIL-spec)
- What movement capability does it have? (Class 25, Class 50—or is it not rated at all?)
- Does the data sheet list your actual substrates and joint geometry?
- What is the service temperature range at the joint location?
- Is the product available in the package sizes your crew will actually use?
Those five questions would have prevented most of the failed assemblies I've audited. The installers aren't the problem. The people who specified the wrong product—or specified nothing at all—are.
So choose foam board for continuous insulation and airtight detailing. Choose fiberglass for cost and breathability. And pair whichever you choose with the right silicone: 3145 RTV when you're bonding, 888 when the joint will move, spray silicone only for maintenance. The data-sheet review is your quality gate. It's far cheaper than the redo.