Radiant Floor Heating in Castle Rock, CO: Decisions Before the Pour

Aug 26

7 min read

If you are building in Castle Rock and radiant floor heat is on the plan, the most consequential decision on the job gets made by whoever ties the tube down the morning of the pour. Not the boiler. Not the thermostat. The height the tube sits at inside the slab, and whether it stays there while the concrete goes in.

Everything upstream of the concrete truck is adjustable later. Boilers get replaced, controls get rewired, circulators get swapped. But tube depth, tube spacing, loop layout, and under-slab insulation are poured in place and stay that way for the life of the building. Tube tied to the bottom of a four-inch pour instead of mid-height raises the water temperature that floor needs by roughly 7°F, permanently, and no equipment you buy afterward buys that back. Castle Rock is still building, which means a large share of the hydronic conversations here could happen while the house is open — the best possible moment, and the one almost nobody uses well.

Get those four right and a mediocre boiler still heats the house. Get them wrong and no equipment purchase fixes it — you run hotter water into a compromised floor and pay the difference every month.

Why does tube depth raise the water temperature for the life of the house?

Concrete conducts heat, but not for free. Heat leaving the tube spreads in every direction, and the more concrete above it, the more goes sideways and down instead of up. To land the same output at the floor surface you raise supply temperature to compensate. Bottom of a four-inch pour versus mid-height is worth about 7°F. It sounds small. It isn't.

It isn't, because that 7°F comes off the top of your margin — and margin is what a modern Castle Rock house should have plenty of. Newer tight construction on the Front Range runs around 25 BTU/hr/ft² of design heat loss, a low number, which is the whole reason radiant works so well in these houses. Low load means low required water temperature, and low water temperature is where condensing equipment is efficient and where the floor stays comfortable rather than hot. Giving 7°F of that back to save an hour of tie-down labor is the worst trade on the job.

What do chairs at 18 to 24 inches have to do with anything?

Chairs hold the tube at the height the design calls for. They go in at 18 to 24 inches along the tube, and the interval is that tight because concrete is heavy and PEX floats.

Under-supported tube does one of two things during a pour. It floats up and sits too close to the surface — striping, uneven surface temperature you can feel through a bare foot, and at worst tube shallow enough to be at risk from a future saw cut. Or it gets stepped on and pushed down, and part of the loop carries the bottom-of-slab penalty while the rest sits at mid-height, so the whole loop needs the hotter temperature.

Nobody sees this happen. It's under four inches of concrete by lunchtime. Which is why it gets checked before the truck arrives rather than argued about afterward.

Can I skip the under-slab insulation to save money?

On a heated slab, no. A slab with tube in it and nothing under it is a large flat heat exchanger with the ground on one side, and the ground never gets satisfied. You are heating the dirt for the life of the house, and the loss is largest exactly when you need output most.

This is the line item that gets value-engineered out because it is invisible and early, and it is the one that costs the most over time. Perimeter insulation matters more than people expect, too — the slab edge is where heat leaves fastest, and on the Palmer Divide the edge is what the wind is working on. Castle Rock sits at roughly 6,200 feet, about 900 feet above Denver, on exposed ground between the metro and Colorado Springs that takes weather the metro north of it does not get. Wind drives infiltration and strips heat off a bare slab edge. Insulate it.

How long should a loop be, and where should the loops go?

Loop length is a balance. Too long and the water gives up too much heat before the circuit closes, so the far end runs cooler than the near end and the room heats unevenly. Too short and you're paying for manifold ports you didn't need.

Layout is a design decision, not something worked out with a chalk line on pour day. The loop under the kitchen island is the clearest example: cabinets and appliances sit on the slab and block heat from reaching the room, so tube under a cabinet run does nothing while still consuming loop length and pressure drop. Same for a pantry or a built-in bench. Those get planned around — which means the millwork drawing matters to the mechanical design, and it matters before the pour.

Bathrooms and entries usually get tighter spacing, as do rooms with a lot of glass, near the glazing. Spacing is how a radiant floor answers a room's actual heat loss instead of an average.

Why do the loops go on 100 psi air and stay there through the pour?

Because a shovel strike should be obvious within seconds, not in December.

Every loop goes on 100 psi of air, held through the pour, gauge where everyone on site can see it. Air, not water, so a leak doesn't soak the sub-base and nothing freezes. If a loop gets nicked by a rebar tie, a stake, or the edge of a screed, the gauge drops and the crew knows immediately — while the concrete is still workable and the tube still reachable. It documents itself, too: anybody walking the slab can see the system was intact at the pour.

Snowmelt works the same way for the same reason — tubing before the concrete, loops on air through the pour, under-slab insulation and 6x6 W1.4 mesh in base scope. If a heated driveway or apron is anywhere in the conversation, it belongs in this same planning window; snowmelt is not something you add to a finished driveway, and summer schedules easiest.

What's different about a staple-up or plate retrofit?

A slab job and a retrofit into an existing floor are different animals. In a slab the concrete is the emitter, the mass is enormous, and the system responds slowly and steadily. In a staple-up or plate system under a wood floor, output climbs through subfloor, underlayment, sometimes carpet — each layer resisting heat and pushing the required water temperature up.

Plates matter here. Aluminum transfer plates spread heat along the underside of the subfloor instead of letting a bare tube radiate into a joist bay. A plated system runs meaningfully cooler water for the same output — which keeps a condensing boiler condensing, and keeps the floor from running hot stripes over the tube with cool strips between. Heavy carpet and pad can eat most of the output you designed for.

Either way the design starts the same place: a room-by-room heat loss on the house as it will actually be built. That's true for radiant and for the boiler feeding it.

What has to be decided before the concrete truck is scheduled?

Short list, all decidable weeks ahead:

• Room-by-room heat loss for the finished house, with the real window package and insulation values.

• Design supply water temperature, which the heat loss and floor construction jointly set.

• Tube depth, spacing, and chair interval, in writing.

• Loop layout with cabinet runs, islands, and fixed built-ins drawn on it.

• Loop lengths and manifold locations — somewhere reachable in ten years.

• Under-slab and perimeter insulation, specified and not negotiable.

• Floor coverings by room, because they change the required temperature.

• Pressure test protocol, and who watches the gauge.

Permits and inspections go through the Town of Castle Rock.

Why does sizing off the plan set go wrong here?

Because the plan set is not the house. It's the house as drawn at bid time, and in a town building this fast the finished house has usually moved. The buyer upgrades the window package. A bonus room gets added over the garage. The insulation contractor beats spec. A wall of glass gets bigger because the view was worth it. Each changes the heat loss, and window upgrades change it twice — a better window loses less heat, lowering the load, and it runs a warmer inside surface, reducing the cold downdraft the floor near that glass was designed to counter. Design to the old plan and you get a floor that overshoots near the glass and a boiler with more capacity than the house can use. (Most boilers replaced in this metro run two to three times larger than the building needs, for exactly this kind of reason.)

The fix isn't complicated: run the heat loss on the house actually getting built, and re-run it if the envelope changes before the pour. Combustion equipment also loses capacity as air thins, and at roughly 6,200 feet that correction isn't optional — manufacturers publish altitude derate tables, and the derate gets applied once. Not zero times, not twice.

Why doesn't a correctly designed radiant floor feel warm?

This is the complaint we hear more than any actual defect, so here it is plainly: radiant floors are not warm floors. They are floors that heat a room.

Floor surface temperature is set by how much heat the room is losing. A well-built Castle Rock house with a low heat load needs very little added to hold temperature, so the floor only runs a few degrees above room temperature — which reads as neutral to a bare foot, not warm. That is the system working, not failing.

Chasing a warm-feeling floor in a tight house means overheating the room, and then you have a house that's too hot and floors that feel about right. If warm-underfoot in one bathroom is genuinely the goal, that's a design conversation up front — usually a dedicated loop with its own control. It is not a reason to raise water temperature across the whole house. There's more on what radiant actually feels like here.

Where do you work?

ByDesign Hydronics & Air works out of Littleton and covers Castle Rock and Douglas County plus the south metro — Highlands Ranch, Centennial, Littleton and nearby. Hydronics only: boilers, radiant floor, snowmelt, mechanical room design. That's the whole list, deliberately.

If you're framing now, the useful call happens before the concrete is scheduled. If you've already poured and something isn't behaving, we'll tell you what the slab is actually capable of and what it would take to work with it, without a running commentary on whoever was there before. If the answer is "leave it," we'll tell you that too.

Call (303) 908-5658 or get in touch here.