As you read this, the Northern United States is in the middle of a cold and snowy winter. With spring several weeks away, many of you are already tired of cold-cranking your car's engine to life after each frigid night. Perhaps you're also sick of chopping ice off the garage floor and walking through the slop your vehicles carry into the garage after every snowfall.

When we built our home in 1980, the solar/energy conservation movement was soaring. The notion of heating a garage occupied by inanimate vehicles was heresy to someone convinced that superinsulated construction and solar energy were the only home-heating concepts worthy of attention.

Been There, Done That, Don't Want To Go Back

For 20 winters we walked out of our warm house to frozen vehicles covered with the same snow, ice and slush they carried in the day before. By mid-winter the floor was either perpetually wet or covered with frozen car droppings depending on how much salt the highway crew spread the day before.

Needless to say, we weren't exactly proud of our garage's appearance this time of year.

As each winter dragged on, I kept asking myself if putting up with the cold, sloppy garage was worth the energy I was saving by not heating it. Holding me back in part was the geothermal heat pump that heated our home at the time. It had no reserve capacity that might be directed into the garage.

Then, in 2001, we replaced the heat pump with a boiler that became the universal “heat server” for all our thermal needs: house, DHW, swimming pool, and yes, even the garage.

My son and I demolished the old floor slab; piped in new floor drains; insulated the walls, ceiling and subgrade; replaced the old overhead doors with foam-core units; and placed two tubing circuits for heating the new slab.

The garage heating subsystem we installed is similar to that shown in Figure 1.

The subsystem (shown in the cloud in Figure 1) connects to the boiler's primary loop with a pair of closely spaced tees. This standard injection mixing piping detail prevents the pressure distribution in the primary loop from influencing flow through the injection risers.

The injection risers connect to the hot-side ports of a small brazed-plate heat exchanger. The injection pump is equipped with an integral check valve. It's operated by a variable-speed injection controller that monitors the outlet temperature from the heat exchanger (e.g., the temperature of the glycol solution supplied to the floor circuits).

The injection controller also monitors boiler inlet temperature. This is necessary because a cold garage slab can place a high transient load on the heat exchanger. High enough to pull the boiler temperature down into the flue gas condensing range, especially if the heat exchanger is a bit oversized.

By monitoring the boiler inlet temperature, the injection controller prevents the heat exchanger from sucking heat out of the primary loop faster than the boiler can produce it.

The integral check valve in the injection pump is to stop migration of hot primary loop water through the heat exchanger when garage heating is off. This prevents repeated heating of the glycol solution within the heat exchanger when the primary loop is operating for other loads.

Elevated temperatures are known to accelerate the breakdown of inhibited glycol into a more acidic solution. Why unnecessarily “cook” the glycol solution within the heat exchanger due to unchecked heat migration?

Some months earlier I had experimented with a spring-loaded brass check valve within a variable-speed injection mixing system to see what would happen. The resulting injection flow was erratic due to the valve's forward opening resistance. The metal disc in the check valve also clicked and clacked as flow passed through in pulses. After a couple days of this, I decided this detail was not going to work and removed the check valve.

When circulators with integral checks first became available, I was skeptical about using them as variable-speed injection pumps given my earlier experience. Still, I decided to try one of these pumps in my new garage heating system, and was pleasantly surprised. There was no noise from the built-in thermoplastic check valve, even at low injection flow rates. When the garage heating is turned on, the injection pump does ramp up to 60 percent to 80 percent speed before the integral check valve pops open.

However, within a few minutes of startup, the supply temperature to the garage circuits consistently settles to the target setpoint of the injection controller. I've watched this subsystem go through its start-up routine many times with different settings for the mixed supply temperature. Each time the system quickly and quietly settled in to the set supply temperature. Based on this, I'm convinced that circulators with integral check valves are acceptable for variable-speed injection mixing applications.

The wiring for the subsystem is shown in Figure 2. Upon a call for heat from the garage thermostat, the garage heating subsystem is turned on by a line voltage output from a standard multizone relay center. Line voltage powers up the 120/24 VAC transformer, which then “wakes up” the injection mixing controller and allows it to take control of injection pump speed.

The strap-on aquastat attached to the water outlet side of the heat exchanger prevents the distribution circulator from operating until heated water is flowing through the heat exchanger. To see why this is necessary, think about what might happen if garage heating was turned on with the slab at a very low temperature. Imagine also that the boiler is just starting up from room temperature, and that the injection pump is off while the controller waits for the boiler inlet temperature to rise.

With frigid glycol racing through a heat exchanger containing a coffee cup's worth of nonflowing water, there's a good possibility the water side could freeze solid before sufficient heat makes its way up from the boiler. The aquastat prevents this by verifying that the water leaving the heat exchanger has reached at least 100 degrees F before allowing the onslaught of cold glycol. Set the aquastat for a wide differential of, say, 30 degrees F to prevent the distribution circulator from cycling on and off.

I also suggest oversizing the expansion tank on the glycol side of the heat exchanger. Going one tank size larger than minimum provides pressurized storage for additional fluid to make up for vented air when the system is first put in operation.

A couple of other slab construction details are also worth mentioning. First, if you have any influence on placement of the floor drains, have one drain installed under the “footprint” of each vehicle. Be sure the floor slopes at least 1/8 inch per foot toward the drain in all directions. This lets the water and slush run under the vehicles on its way to the drain rather than into the walkways around the vehicles. With this detail, the floor will seldom be wet more than a few inches out from the vehicle.

Secondly, don't insulate the edge of the garage slab at the overhead doors. This allows the snow and ice to melt a few inches out from the door. It also prevents the door's bottom gasket from freezing to the slab. Your garage door opener will appreciate this detail when it's time to lift the gasket off the floor. All other edges of the slab as well as its underside should be insulated. We used 1 1/2-inch thick/25 psi-rated extruded polystyrene, which has plenty of bearing strength for cars and pickups.

Finally, be sure the exposed walls and ceiling of the garage include a continuous vapor barrier. If the garage is extremely tight, you might also want to include a small fan or other means of ventilation to control humidity.

Don't Wait 20 Years To Do This

We now turn on garage heating when the weather is such that the garage won't maintain 45 degrees F. Vehicles that entered covered with slush the previous evening are dry the next morning as is the floor around them. It's satisfying to hear “liquid snow” running into the drain sumps under each car within a few minutes of pulling in and closing the overhead door.

I've estimated that operating our garage this way requires about 25 gallons of No. 2 oil per season. Currently that costs us about $30. At this point in life, I've concluded that's a small price to pay for getting into a dry and reasonably warm vehicle during those long northern winters.

I now emphasize that garage heating is a “must have” when planning a new hydronic space heating system in a cold and snowy climate. When a customer hesitates based on added cost, I strive to convince them to at least install underslab insulation and tubing circuits before pouring the slab. It's a safe bet they'll want their “radiant ready” garage floor to begin radiating before too long.