Building a Heated Subfloor for a 4-Season Transit Van

The Final Heated Subfloor Stack

From bottom to top, the subfloor assembly will be:

The heated floor is not intended to be the van’s only heat source. In a Saskatchewan winter, the Rixen hydronic system and fan coils still need to do the serious heating work.

The floor heat is about comfort. It should take the edge off the cold floor, make the living space feel more even, and avoid that brutal “standing on a frozen metal box” feeling that can make a winter van feel miserable even when the air temperature is technically warm.

Why I Chose Rixen for Heat

I went with the Rixen MCS7 gasoline hydronic system because I wanted an integrated heat and hot-water system that fits the way I plan to use the van.

The van is a gasoline Transit, so the gasoline-fired hydronic system keeps the fuel source simple. The complete kit gives me the foundation for cabin heat and domestic hot water. The floor heat kit adds the radiant comfort layer. The engine waste heat kit allows the system to capture heat from the engine coolant while driving, which should reduce the need to burn fuel for heat or hot water after travel days.

That matters for the kind of use I’m planning: cold-weather travel, shoulder-season camping, and real winter capability without building a system that feels cobbled together from unrelated parts.

I also like that this gives me a repeatable platform. If I eventually build vans for customers, a well-documented hydronic system is easier to explain and support than a completely custom collection of parts.

Why SOPRA-XPS 30 Instead of Coosa Board

Early in the floor planning, I looked seriously at Coosa Bluewater board.

Coosa is an appealing material. It is a fiberglass-reinforced polyurethane foam panel, commonly used in marine applications where builders want something lighter and more rot-resistant than plywood. It has good structural qualities, handles moisture well, and feels like the kind of premium material that belongs in a high-end build.

The problem is that Coosa solves the wrong problem for this specific floor.

For a Saskatchewan-capable van floor, insulation matters more than composite bragging rights. Coosa is structurally impressive, but it does not insulate nearly as well as XPS. A 1″ Coosa panel is roughly in the R-2.5 range, while 1″ XPS is around R-5. That is a big difference when the floor is sitting above cold steel in a prairie winter.

Cost was the other issue. I was able to get Coosa Bluewater 20 delivered to Saskatoon, but by the time I priced the sheets and freight from Vancouver, using it across the full cargo floor did not make sense. An extended Transit would realistically need around three sheets for the floor area, and that would put the material cost far higher than I could justify for a layer that still underperformed XPS thermally.

So I chose SOPRA-XPS 30.

• Approximately R-5 per inch
• Closed-cell insulation
• Better availability
• Lower cost
• Easy cutting and routing
• Enough compressive strength when paired with a proper plywood skin
• A more repeatable path for future builds

Could I have chased 40 psi or 60 psi XPS? Yes. But in Saskatoon, the higher-density commercial products quickly become harder to buy in small quantities. Some suppliers will sell them by the pallet, but not by the sheet.

SOPRA-XPS 30 became the best practical balance: available, insulating, workable, and strong enough when used as part of the full floor sandwich.

Filling the Transit Floor Ribs

The Transit cargo floor is corrugated. If I laid the 1″ XPS directly over the rib peaks, there would be voids below parts of the foam layer. That could lead to movement, squeaks, less support under the routed PEX, and uneven contact across the floor.

To avoid that, I’m cutting 3/8″ strips from 3″ SOPRA-XPS 30 board and using them as rib fillers between the corrugations.

The rib fillers are not structural sleepers. They are there to:

• Reduce voids
• Support the main foam layer more evenly
• Reduce air movement under the floor
• Reduce rubbing and squeaks
• Maintain the insulation strategy with closed-cell material
• Avoid using wood directly down in the metal floor valleys

I do not want moisture-wicking wood strips sitting in the lowest parts of the steel floor. Foam rib fillers are lighter, warmer, and less likely to create moisture problems.

Routing 3/8″ PEX Into the XPS

The hydronic floor loop will use 3/8″ PEX routed into the 1″ SOPRA-XPS 30.

I chose 3/8″ instead of 1/2″ because this is a thin van floor assembly. Smaller tubing gives me a shallower route, tighter layout options, and more insulation remaining below the tubing.

That is important because every fraction of an inch matters here. I want enough insulation below the tubing to reduce heat loss into the steel floor, but I also need the tubing close enough to the top of the foam to transfer heat into the aluminum sheet above it.

The goal is for the PEX to sit very close to the top surface of the foam, ideally just proud enough that the aluminum layer and plywood above help hold it in contact. It should not be crushed or pinched, but it also should not sit low in the groove with an air gap above it.

Good contact matters more than adding messy thermal paste or trying to overcomplicate the system.

Why I’m Using Full-Sheet Aluminum Instead of Separate Heat Plates

A lot of hydronic radiant systems use aluminum heat transfer plates. I considered that route, but with this floor I’m using a continuous 1/16″ aluminum sheet above the routed PEX instead.

The aluminum sheet becomes the heat spreader.

In a van floor, full-sheet aluminum has a few advantages:

• More even heat distribution
• Fewer individual parts
• Fewer opportunities for plates to rattle or shift
• Easier repeatability once the floor template is established
• Cleaner coverage over the entire routed tubing field
• Added protection over the PEX channels

I also considered 1/8″ aluminum, but that adds a surprising amount of weight. Across the floor of an extended Transit, the difference between 1/16″ and 1/8″ aluminum is not trivial. The thicker sheet would add stiffness and thermal mass, but I do not think the benefit is worth the extra weight.

A 1/16″ aluminum sheet is already substantial enough to do the heat-spreading job in this assembly, especially with 3/8″ PEX routed directly below it and 12mm plywood above it.

I am not planning to use separate heat transfer plates unless testing shows obvious hot and cold striping. My expectation is that the full aluminum sheet will be sufficient.

Keeping the Aluminum Isolated From the Van Steel

One detail I do not want to miss is keeping the aluminum sheet from directly touching the van’s steel floor.

That is mostly about galvanic corrosion. When two different metals touch each other and moisture is present, they can create a small electrical reaction. Add winter road salt, condensation, wet boots, freeze-thaw cycles, and hidden seams, and the risk becomes more relevant in a Saskatchewan van than it might be in a fair-weather build.

The goal is not to be paranoid. The goal is simply to avoid building aluminum-to-steel contact points into a floor assembly that will be difficult to inspect later.

In this design, the aluminum sheet sits above the 1″ SOPRA-XPS 30 layer, so it should already be isolated from the steel floor by the foam. The places I need to be more careful are the edges and penetrations:

• Door thresholds
• Seams between aluminum sheets
• Wheel well transitions
• Tiedown or cabinet anchor points
• Any fasteners that pass through the aluminum layer
• Any exposed aluminum edges near the steel body

Where needed, I’ll use the foam itself, sealant, butyl, paint, plastic washers, or other non-conductive breaks so aluminum is not sitting directly against bare steel. I also want to avoid metal-on-metal rubbing, which can create squeaks, ticking, or fretting as the van flexes and the floor assembly expands and contracts.

The practical rule is simple: the aluminum can touch foam, plywood, adhesive, sealant, or finish materials. It should not directly touch bare van steel.

Good Contact Matters More Than Extra Complication

The heat-spreading layer only works if heat can actually get into it.

That means the PEX routing matters. I want the tubing to sit close to the top of the routed foam, with the aluminum sheet held snugly above it by the plywood layer. The tubing should not be crushed or pinched, but it also should not sit buried too deeply in the groove with an air gap above it.

For this build, I’m keeping the detail simple: accurate routing, close contact, a continuous aluminum spreader, and a tight floor sandwich.

That approach is cleaner, easier to repeat, and less likely to introduce messy materials or long-term service issues.

Why 12mm BS1088 Marine Plywood

I debated whether the aluminum layer would let me drop to thinner plywood, such as 9mm.

In the end, I chose 12mm BS1088 marine plywood.

That choice adds a little height and weight, but it also adds confidence. The plywood is not just a cover panel. It is the structural walking surface, the load spreader, and the layer that makes the floor feel solid instead of delicate.

The other reason for marine plywood is consistency. A van floor is a rough environment: vibration, seasonal temperature swings, wet boots, condensation risk, plumbing, cabinetry loads, and the possibility of small leaks over time. Regular construction plywood can work in many van builds, but it often comes with more voids, less consistent veneers, and less predictable behaviour when it is cut, fastened, or exposed to moisture.

BS1088 marine plywood is not magic, and it still needs sealed edges and good detailing. But it is made for more demanding moisture-prone environments, with better veneer quality and waterproof glue standards than ordinary sheathing plywood. In this floor, that matters because the plywood is part of the structural sandwich above the routed foam and aluminum layer.

The 12mm marine plywood gives better margin for cabinetry, water tank loads, battery compartment planning, fridge slides, wet bath structure, foot traffic, vibration, future service work, and repeatability in later builds.

The Transit gives me more headroom than many vans, but I still care about floor thickness. I just do not want to save height by building a floor that feels compromised.

The 12mm BS1088 plywood is the right call for this version of the floor: thin enough to preserve headroom, but strong and stable enough to make the heated subfloor feel like a premium build instead of an experiment.

3M 4200 vs. Sikaflex: Why I Chose 3M Marine Adhesive Sealant

This was one of the decisions I went back and forth on.

A lot of van builders use Sikaflex. It is almost the default answer in many van build videos and forums. Sikaflex products are popular for good reason: they are flexible, bond well to many surfaces, and handle vibration better than rigid construction adhesives.

For this floor, though, I chose 3M Marine Adhesive Sealant Fast Cure 4200.

The reason is not that Sikaflex is wrong. It is that 3M 4200 fits the specific role I want this adhesive to play.

I am not using the adhesive as the main structural element of the floor. The floor structure comes from the sandwich: XPS, aluminum, marine plywood, cabinetry, thresholds, and mechanical capture. The adhesive is there to act as a flexible bedding layer and anti-squeak material, especially where foam and rib fillers meet the painted steel floor.

For that job, I want an adhesive/sealant that is flexible after cure, vibration tolerant, water resistant, suitable for marine-style environments, strong enough to hold and bed materials, and not so permanent that service becomes impossible.

That last point is important.

3M 5200 is famous for being extremely permanent. It is an excellent product when you want something to stay bonded almost no matter what, but that is not what I want under a van floor. If I ever need to remove part of this floor, I do not want the adhesive to turn service work into demolition.

3M 4200 is the more moderate choice. It is still a marine adhesive/sealant, but it is generally considered more removable than 5200. That makes it a better fit for bedding and anti-squeak use in a floor assembly.

Compared with Sikaflex 291i, 3M 4200 is not a dramatic departure. Both are flexible marine adhesive/sealants. Either one could probably work in this application if tested properly. The reason I’m comfortable with 4200 is that it gives me the flexible, water-resistant, vibration-tolerant behaviour I want while staying on the more serviceable side of the marine adhesive spectrum.

The key is how it is used.

I am not planning to trowel it across the entire floor. I’ll use controlled beads and dabs where I want bedding, isolation, and squeak reduction. The floor should be captured, not permanently entombed.

Because I’m using XPS, I will still test 3M 4200 on scraps before committing across the van. Polystyrene foam can be sensitive to certain solvents, and I do not want to assume compatibility just because a product works well in marine applications.

The decision is not really “3M is good and Sikaflex is bad.” It is this: Sikaflex would likely work. 3M 4200 should also work. I chose 3M 4200 because it matches the flexible, marine-grade, semi-serviceable bedding role I want in this specific floor assembly.

Why This Floor Is a Compromise I Can Live With

There is no perfect van floor.

A thicker floor would insulate better, but I would lose standing height. A Coosa-based floor would be structurally impressive, but it would cost far more and insulate less effectively. Higher-density XPS would be nice, but it is harder to source locally in small quantities. Thicker aluminum would spread heat and add stiffness, but it would also add unnecessary weight.

This final assembly is not the most exotic version possible. It is the most balanced version for this van.

• R-5 main insulation layer
• Hydronic floor comfort heat
• Low overall floor height
• Closed-cell rib filling
• Full aluminum heat spreading
• 12mm marine plywood stiffness
• Flexible anti-squeak bedding
• Materials I can actually source
• A system that could be repeated in future builds

That repeatability matters.

This van is personal, but it is also part of a larger experiment. I’m using this build to test whether a small, high-quality camper van conversion business could eventually make sense. For that to happen, the systems need to be more than clever. They need to be explainable, serviceable, reliable, and buildable more than once.

That is why I like this floor.

It is not just a pile of premium materials. It is a set of deliberate tradeoffs.