After 30 years, it was time for a solar refresh.

I’ve lived with an array of solar thermal collectors on our house for the last 30 years. Those collectors came from Revere Solar & Architectural Products, a division of the now defunct Revere Copper and Brass, where I worked as an engineer from 1978 to 1981.

Figure 1 shows the collector array. Each of the six collectors measured 35 inches wide by 77 inches long. Those dimensions are 1 inch longer and wider than the glass used in standard sliding patio doors. Because it was mass produced, 34-inch by 76-inch tempered glass was available at lower cost, and thus became the benchmark around which the remainder of the Revere collectors were designed.

Other than size and insulation material, the Revere collectors are similar to current flat-plate collectors. They had all-copper absorber plates with a black chrome “selective surface” on the upper side. The absorber plates were mounted in an extruded aluminum housing, covered with low-iron tempered glass, and insulated with low-binder, high-temperature fiberglass.

Initially, these collectors had thermal performance indices similar to a modern flat-plate collector (FRta = 0.72, and FRUL = 0.83). However, I’m convinced some of this initial performance was lost due to the results of aging on the selective surface, insulation R-value and perhaps even some changes in the transmissivity of the glazing.

Back in the 1970s, engineers at Revere estimated their collectors would have an average design life of 20 years. Apparently I got collectors destined for the right side of the statistical bell curve. They were still going at 30 years of age, albeit with some loss of performance and a few repairs of minor pin-hole leaks. However, I felt it was time for a solar “refresh” on our roof and a graceful retirement of the Revere collectors. At the risk of sounding sentimental, those aging collectors had done their job well and were a reminder of good years when a “wet behind the ears” engineer learned a bit about solar thermal systems.

Waiting In The Wings

It just so happened that Caleffi North America finished testing its new patent-pending “five-port” flat-plate collector at the same time I decided on the solar refresh at our house. This new collector could be configured for either closed loop/antifreeze systems or drainback systems. I’ve long been a fan of the latter, and wanted to keep the system operating in its original drainback configuration.

Figure 2 shows how the Caleffi drainback collector is designed. Its absorber plate has slightly sloping internal headers that form a very shallow “V” shape. Both ends of the upper and lower headers terminate just outside the anodized aluminum enclosure with union-type piping connections. These are the connections one would use if connecting the collectors for a standard antifreeze-type system.

The unique feature is found at the bottom center of each collector. The low point of the bottom header is configured as a “T” to allow all fluid to drain out of the absorber plate. Thus, when used in a drainback system, these collectors would have an external lower header in combination with an internal upper header. It’s a simple and unique idea that eliminates the need to side-slope the collectors, and perfect for the retrofit situation I had.

Knowing my interest in these collectors, my friend “Hot Rod” Rohr, who is Caleffi’s director of training and the inventor of this five-port collector, made me a great offer. He would personally come out and help me install the new collectors. I’ve known Hot Rod for many years. We would see each other at trade shows, conferences and other such events. Having the opportunity to turn the wrenches with this very generous and experienced hydronics artisan was an opportunity I couldn’t pass up.

Before Hot Rod arrived, my wife and I managed to remove the old exposed piping and lower the old collectors off the roof without any problems. Gravity was on our side.

Once Hot Rod arrived with the new collectors and installation tools, the first task was getting the mounting brackets and rails in place. Based on the mounting flange that runs around the lower perimeter of the collectors, we decided to use aluminum Unistrut to support the array. The Unistrut was secured to 1 1/2-inch x 1 1/2-inch x 1/4-inch aluminum clips I fabricated from angle stock. We decided to support these clips on top of small pieces of pressure-treated lumber to help spread the load over the shingles. We snapped chalk lines to maintain alignment of the Unistrut. One rail would be located about a foot below the upper edge of the collector and the other about one foot above the lower edge.

Doubled-sided roof sealing tape was applied to the back side of each treated wood support. This tape has an amazing ability to stretch as can be seen by how it “coats” the threads of the lag screws in Figure 3. It definitely beats dealing with asphalt roofing mastic on a warm and sunny day.

The aluminum clips and their supporting blocks were lagged into pilot holes drilled directly into the upper chords of the roof trusses. These supports were spaced 48 inches apart. We used a small drill bit as a “probe” to find the center of each truss, and then sealed up the probe holes with a dab of silicone caulk. Hot Rod’s cordless impact driver was ideal for driving home the 3/8-inch lags (Figure 4).

The aluminum Unistrut was supplied in 9-foot lengths. Butt ends were joined using stainless-steel splines supplied by Unistrut and held in place with set screws. The splines fit snugly and allow for accurately aligned strong joints. We staggered these joints so that they did not to fall directly over the mounting clips. Spring-loaded stainless-steel T-bolts with 1/4-inch by 20 TPI threaded studs, also supplied by Unistrut, provided the hold-down for the collector mounting clips (Figure 5).

After the Unistrut was in place, we drilled several 1/4-inch holes near the lower edge to drain out rainwater.

Going Up

With the rails in place, it was time to hoist the collectors. The 4-foot by 8-foot panels weigh about 90 pounds each. We set up an extension ladder as a slide and slid each collector as far up as possible while standing on the deck below the roof. My wife, Joyce, along with some help from friction, held each collector in position on the ladder while Hot Rod and I got back up on the roof for the final lift. We discovered that leaving the cardboard packing on the ends of the collector provided protection while sliding them up the ladder.

The first collector was carefully positioned, leveled and secured at all four locations where its edges crossed the Unistrut rails as shown in Figure 6. The mounting clips coordinate with the collector-edge flange, and tightly grip the Unistrut as the nut is tightened on the T-bolts inside the Unistrut.

With the first collector in place, the others quickly followed. Each connected to its neighbor using the double-nutted brass fittings. A watertight seal is created at each connection using a flat gasket that swells upon contact with water. Figure 7 shows how the double-nutted brass fittings go together. The space between adjacent collectors is only 3 inches - just enough for comfortable access to the mounting hardware.

The low-point connection (e.g., the fifth port) on each collector was adapted to 3/4-inch copper tubing. We used a copper street ell on the left collector and cut progressively longer tube stubs for the other four. This allowed the lower manifold to slope at 1/4-inch per foot for proper drainage (Figure 8).

The lower header was insulated with 1/2-inch-thick foam rubber insulation as the header was made up. The end of the header passes directly through the siding at the right side of the recessed array. From this location we were able to maintain a minimum 1/4-inch-per-foot slope back to the mechanical room.

As of this writing, the system is operating on the original differential temperature controller. The collector temperature sensor is temporarily inserted into the sensor opening in one of the collectors. The next step in the renovation is to change to a Caleffi controller, which has a sensor that fits into a well fabricated into the top left corner of each collector. This puts the sensor in direct contact with the absorber plate, an important detail that speeds control response in drainback systems.

With the piping completed, we flipped the system controller on, snugged up a couple collector fittings and cycled the array through a few fill/drain cycles. As has been true during the previous 30 years, gravity didn’t let us down. The system filled and drained as expected (Figure 9).

Hot Rod is looking into the possibility of having one of the original Revere collectors tested for its current thermal performance indices. It would certainly be interesting to know how much efficiency 30 years has chipped away.

Perhaps in another 30 years the next generation of solar professionals will be changing out these collectors with the latest technology. Hot Rod and I will probably be directing work from the ground at that point. Perhaps those new collectors will be constructed of the latest composite materials and placed by a solar-powered construction robot. Whatever the case, the sun will still be out and gravity will still be working.

In closing, a special thanks to my friend Hot Rod Rohr, whose efforts made this project an efficient and enjoyable way to spend a couple sunny spring days.