get quotes get quotes get quotes


By Solarbair, LLC © November 24, 2009


This retrofit system features: drain back ground mounted collectors, large underground storage, solar hot water baseboard and a guilt free hot tub.  It has cut the propane use for my 2,400 square foot home and garage in half and reduced the power usage of my hot tub.  It has also functioned, trouble free through the coldest days of winter and the hottest days of summer.  Let me tell you…

                It all started with a house heated by a cast iron boiler with standard hot water baseboards.  With a finished basement, a garage full of tools and a fondness for a warm car on cold Montana mornings I had to find a place for the hot water storage without giving up a single closet.  I didn’t want to spend $1,000 for every 100 gallons of storage either.  


underground tank

 I called up Anderson pre-cast the local septic tank manufacturer. It turns out they can cast a two compartment 1,500 gallon concrete tank with custom openings  for something like $1,200 delivered.  They will even haul it up and set the 15,000 lb tank in the hole.  That is 1,000 gallons of storage with the 500 gallon dry compartment as a pump room, sauna and/or bomb shelter.  More space, not less!  I had some concern that concrete could not take the heat.  But if the wet side is lined with a pond liner or high temp EPDM rubber then it won’t leak even if the reinforced concrete does crack.   So far the tank has not been a problem.  In fact, all that concrete provides even more thermal storage.  With 9,288 pounds of concrete heating up and cooling down around that 1,000 gallons of storage every degree Fahrenheit amounts to 1,950 BTU (tank) and 8,330 (water) for a total of 10,280 BTU per degree. In my usable temperature range of 127˚F to 180˚F, tank and water store 544,840 BTUs.  Enough to keep us warm and in domestic hot water through 2-3 cloudy winter days.    Thought of another way; that half a million BTUs is about 7 hours of run time on the cast iron boiler or 6.7 gallons of propane.

I found some high temperature foam insulation (Atlas R-board, R-6 per inch) and surrounded the tank with 4 to 6 inches (more on top less on the bottom).  Interestingly the heavy tank did not damage the insulation underneath.  Then I epoxied some fiberglass risers over the 24” diameter manholes, ran some insulated 4” PVC sleeves below frost to the hot tub and the house and buried it all.  The top of the tank ended up being about 30” below ground level.  Doing the calculation to see how well the tank would hold heat, I came up with stand-by losses of approximately 0.1 ˚F per hour or 2.4 ˚F per day with the tank at 170˚F and an air temperature of 0˚F.  Actual tank losses average about 2˚F per day.  Compare this to a standard solar storage tank which might have stand-by losses of 12˚F per day and this buried tank is looking better all the time.


The collectors needed to be ground mounted near and above the buried tank since I am saving my roof for PV panels. I toyed with building a steel and/or concrete mounting framework. But the costs associated with ready mix concrete and a pump truck for the piers or heavy steel I-beams and welding were more than I wanted to spend.  A steel framework can also act as a ground for lightning which would not be good.  An ingenious fellow at the lumber store recommended 6” x 6” posts treated for direct contact with the ground.  For the horizontal members he recommended the use of the same treated 6” x 6” posts sandwiched between treated 2” x 10”s.  Building it this way eliminates the need for larger dimensional lumber (prices go up fast for larger sizes) and it is all pretreated so there is no sealing or painting.  As for durability it is similar to pole barn construction where treated posts are used in lieu of a traditional foundation.  Since banks will mortgage a pole barn, I am optimistic the posts will remain solid for years to come.  Because of the strong winds we get, I buried the posts at least 5’ deep.  For additional staying power, I ran a couple of pieces of rebar through the bottom of each post and chucked in two 80 lb. bags of concrete with a few rocks as an anchor.  The horizontal beams are sloped 3/8” per foot to facilitate drainage.  This construction along with the Sun Earth aluminum mounting system and a few diagonal braces for additional support has held the collectors in place through winds which rattle the house and can blow a table or a barbeque off a deck. 

ground mount collectors

I decided on eight 4’ x 10’ Sun Earth EC (black chrome selective coating) collectors.  The SRCC efficiency ratings for these collectors are excellent.  They are reasonably priced and are certified for use with drain-back systems.   The larger 4 x 10 collectors also cost less per sq. ft. than smaller collectors especially when mounting costs are factored in.  The reason I stopped at eight collectors (besides the cost) is that Sun Earth recommends not more than eight collectors be piped in series.  More collectors and I would have to pipe the collectors as two banks in parallel, which would complicate the piping.  The eight collectors yielded a gross collector area of 320 sq. ft. and a net area of 298 sq. ft.  Our latitude here in Montana is about 45˚ degrees, for space heating a collector tilt of latitude plus 15˚ is recommended, so the collectors are at a 60˚ angle.  This steep angle helps to avoid over heating in the summer, collects heat nicely in the winter and sheds snow very well.  For space heating systems 1.75 to 3 gallons of storage per sq. ft. of collector is typical, so my storage is a little oversized.  This gave me some concern at the design stage but has caused no problems in practice.  My current feeling is more storage is better than less.  More storage acts as a heat reservoir providing heat longer through the nights and cloudy times. Larger storage also reduces instances of high limiting when heat cannot be collected because the storage tank is at its maximum temperature.  The weather often comes in runs of a few sunny or stormy days in a row.  During a string of sunny days a larger storage tank can capture more heat, when the clouds roll in that heat is available through the first overcast days.  

The water in the solar storage tank is circulated from the bottom of the tank, through the collectors and back into the top of the tank when there is heat to be collected (see pink loop in diagram).  A well placed flow meter acts as a site glass for monitoring the water level in the tank.  Whenever there is no heat to be collected or if the power is out, the water drains out of the collectors and back to the tank because the piping is graded.  There can be no traps in the piping and the end where the water drains back into the top of the tank must not be submerged.  Air has to be allowed into the piping to displace the water.  This is the nature of a drain-back system; freeze protection by gravity.  The system has withstood -40˚F nights and produced hot water on a 0˚F sunny day without difficulty.  Sun Earth drain-back systems are OG-300 certified for locations with record lows down to -50˚F.  Snow on the ground is actually a good thing, reflecting the sun onto the collectors for increased performance.    

piping diagram

A little information on how the drain back system functions might be of interest.  With the system running, tank temperature increases by a little more than 1˚F every fifteen minutes.  A good solar day can raise the tank temperature by 30˚F.  Drain back systems have a 10-15% efficiency advantage over anti-freeze systems. When the collectors heat up and the pump turns on it takes 1 minute to fill the collectors and another 15 seconds to establish a siphon where the entire loop is full of water.  As the collectors cool and the pump turns off it takes 35 seconds to drain the supply line from the top of the collectors back to the top of the tank.  It then takes another 110 seconds to drain the collectors and return piping back through the pump.  Good insulation on the lines to and from the collectors is critical.


The hot tub loop is shown in green on the diagram.  The heat exchanger (HEX) in the hot tub consists of twenty feet of ½” L copper run around the walls near the bottom of the tub.  I used plastic talon nail straps (without the nails) to hold the pipe ¼” off the walls of the tub.  The HEX located near the top of the storage tank is a 25’ coil of ½” K copper.  This loop is filled with propylene glycol anti-freeze since the piping could freeze where it loops over the rim of the hot tub. 

hot tub heat exchanger

                This set up can heat the hot tub from well water temperatures to 104˚F in less than 24 hours and maintain that temperature without difficulty.  However, when the cover is off the hot tub heat-loss often exceeds the heat gain provided.   I have had no problems with the piping discoloring or melting the tub.  Even with the solar storage tank near its high limit temperature of 180˚F the HEX in the hot tub does not cause burns when touched.  At more intermediate temperatures it even acts as a nice foot warmer. 

                When I conceived of the solar hot tub heater I imagined I would set the solar heater to 104˚F and the electric heater inside the tub to 102˚F.  This way the electric heater would only come on when the solar heat had been depleted.  In practice this has not been the case.  Apparently there is an over temperature safety switch in the circuitry of the electric heating system which trips when the hot tub is heated above the electric heaters set point.  The solar heats the hot tub above the electric heaters setting and the electric heater trips out until the tub has cooled and is manually reset.  I have contacted the hot tub manufacturer but as yet have not resolved this problem.  My current solution is to heat the hot tub with solar during the summer and use the electric heater in the winter when all the solar heat can be used for space heating.  This problem may be peculiar to my brand of hot tub or it may be common to several brands.  These kinds of things can happen when we travel off the beaten path.


Because the existing heat system is under pressure we cannot circulate the unpressurized water in the solar storage tank through the heat system.  A heat exchanger (HEX) is required to transfer the heat from the storage tank to the heating system.  The HEX in this case is constructed of a 1” copper line branching into three 25’ coils of ½” K copper run in parallel and then converging back to 1” copper piping.  I positioned the HEX near the top of the tank so it would be submerged in the hottest water.  The performance of this HEX is difficult to quantify.  What I can tell you is this: when the water at the top of the tank is 128˚F or more, the indirect water heater produces water upwards of 120˚F.  To accomplish this, heat must first be exchanged from the solar tank to the heating system and then exchanged again (in this case through a double wall heat exchanger in the indirect water heater) to the domestic hot water.  Suffice it to say heat has been transferred without disappointment.   A note of caution, if there were a leak in the HEX, water would flow from the heat system into the solar storage tank.  This could raise the water level in the tank above the supply pipe from the collectors which could be catastrophic.  The collector might fill and then be unable to drain during the freezing night.  For this reason I have shut off the fill valve to the heat system, limiting the amount of water which could leak into the storage tank.  An over flow could also be installed to offer similar protection.

 Connecting the main heat exchanger to the heating system was easier than I expected.  All I had to do was install a check valve after pump P1 and connect the heat exchanger loop including pump P2 as shown to the right of the baseboard supply and returns.  The system can automatically switch between pulling heat from the storage tank and operating the boiler by powering P1 or P2.  Another advantage of this design is that the solar heated water is not circulated though the boiler.  P2 can be pumping 150˚F water though the baseboards or indirect water heater while the temperature gauge on the boiler reads 80˚F.  We are not using solar energy to heat up 300 pounds of cast iron unnecessarily.

It is common knowledge among heating professionals that fin tube baseboard heaters require very hot water (about 175˚F) to function.  This is not exactly true.  If you want to get 500 - 600 Btu/hr out of every foot of baseboard you do need very hot water.   But baseboards do not cease to put out heat as soon as the supply drops below a certain temperature; the heat output simply decreases.  For instance, my baseboards are rated to put out 250 Btu/hr per foot when supplied with 130˚F water.  Since they were sized (oversized) by rule of thumb they have functioned with supply temperatures 128˚F and above through the winter.  The reduced output due to the lower supply temperatures has been compensated for by longer run times and slower recovery from night time setback.

Every system needs a brain and this one is no exception.  The collector pump, P3 in the diagram, is operated by temperature differential control.  This control has a temperature sensor in the bottom of the storage tank and one near the top outlet of the collector array.  When the collectors are 15˚F hotter than the bottom of the tank the pump is turned on.  If the collector temperature drops to 3˚F warmer than the tank the pump is switched off.  I installed an additional set point control as a safety.  This control has a sensor located in the array and will not allow the collector pump to function unless the collectors are above 110˚F.  This eliminates the possibility of freezing due to a malfunction or improper parameter setting in the differential control.   The differential control logs the highest and lowest temperatures reached and overall heat collected.  It can also shut down the pump when maximum storage temperature is reached.

                There is a second set point control with a sensor near the top of the storage tank.  This control determines if there is usable heat in the storage tank.  If the top of the tank is 128˚F or more pump P2 will be powered upon a call for heat from the thermostats or indirect water heater. When the tank is below 128˚F pump P1 is used.  There is a third set point control located near the hot tub which maintains the hot tub temperature at 104˚F.  If the hot tub drops below 104˚F and the storage tank is over 128˚F, pump P4 is activated until the hot tub is back up to temperature.  All of the temperature settings can be adjusted to improve efficiency and comfort.


This solar heating system has cut my propane usage by about 50%.  Down from about 900 gallons a year to between 400 and 500 gallons.  At the current price of $1.98/gallon for propane there is a savings of around $891 per year.  There are also some electrical savings.  The solar system heats the hot tub for about five months out of the year.   It also contributes heat to the built in garage offsetting the electrical space heater there.  I would estimate the total cost savings at over $1,000 per year.  Carbon dioxide emissions are reduced by 2.9 tons which is equivalent to .5 cars or light trucks taken off the road. 

The entire system cost us right around $18,000, we did almost all the labor, paying only for some backhoe work and the materials.  We anticipate a 30% federal tax credit and $1,000 from the state of Montana which means we will have spent $11,600.  Simple pay back is, well, simple at 11.6 years.

To illustrate how such a system might be leveraged, consider the following:  Borrow the entire $18,000 at 5% interest for 20 years for a yearly payment of $1,444.  The first year you get tax credits of $6,400, save $1,000 in energy costs and pay $1,444 on the loan to come out $5,956 ahead.  Over the next 12 years the loan will cost more than the energy saved and the $5,956 will be whittled down to about $3,700.  This is the low point of the payback and you are still $3,700 ahead because you installed the solar heating system.  During the years from 12 to 20 your savings will climb back up to about $5,200.  From that point on any energy costs saved go straight into your pocket.

payback with loan

If you do what we did and pay for the system up front (no loan) the savings are not as immediate but larger in the long run.  We start out the first year by paying $11,600 (after tax credits) and saving $1,000 in energy costs.  After 10 years we have recovered our investment and at 20 years we are $17,000 ahead.  Having saved the interest paid on the loan in the previous example.  

payback without loan

There is also an equity increase in the home.  As soon as the solar project is up and running the value of the property is increased.  Who would not pay more for a home with a functioning renewable energy system installed on it? According to my insurance agent, my solar system is covered, at no additional cost, by my home-owners policy.  Also, keep in mind that home energy costs are usually paid with after tax dollars, while interest on a home mortgage can be paid with pre-tax dollars. If you are building a home or refinancing you may be able to add the cost of the solar system into the mortgage to reap additional tax savings. 

To be fair I should mention that there will be maintenance costs on the solar system over the next 20 years.  These maintenance costs can be offset in at least two ways.  First, if the system is maintained it will continue to function beyond the 20 years depicted and provide additional savings until it is decommissioned.  Second, whenever the solar system is satisfying a heat load the back-up heat source is idle, so that maintenance or replacement of the boiler or hot tub heater is delayed.

The payback for a solar space heating system has one big strike against it right off the bat.  There is more solar energy available in the summer and more space heating demand in the winter.  In order to provide space heating in the winter the system must be oversized for the summer load.  We paid for a system big enough to partially heat our house in the winter but we cannot use all of the heat produced in the summer.  We could heat several hot tubs and our domestic hot water (DHW) no problem during the warm sunny months.  The free labor we invested in the system makes up for this.  Because of our labor investment the system has payback figures comparable to those seen for professionally installed DHW solar systems.   Residential solar systems which heat DHW can be sized so that almost all the solar energy collected can be put to use year round.  The payback on such systems is about as good as it gets for renewable energy.  The simple payback is often under 10 years even in cold climates and substantially better in warm sunny regions.  Some of the best paybacks for solar thermal systems can be realized in commercial situations by hotels, car washes and laundry mats where large year round hot water demands coincide with available solar energy.  


                I knew from the outset that if our house were heated with high mass radiant floors the solar advantage would be better.  The solar tank would contain usable heat down to 100˚F or perhaps even lower depending on floor coverings and heat loss.  This would increase usable heat storage by a whopping 60%.  Collector efficiency at the lowest usable temperature would increase from 55% to 62%, resulting in almost 13% more BTU’s collected.  The combination of these two improvements would shorten the payback period substantially.  In retrofit applications, like mine, similar advantages can be had by installing larger or more radiators.  If ever I build again, solar radiant heat will be in the plans from the beginning. 

Further efficiency gains could be had if the domestic hot water feed line ran through a heat exchanger in the solar storage tank before entering the water heater.  In this way DHW would be preheated as long as the solar storage tank was hotter than the incoming well water.  If the HEX was large enough and my back-up water heater was an on demand type no further DHW heating would be required.  

 In the past year variable speed pumps have made substantial inroads into the hydronics market.  If I had it to do over, I would incorporate variable speed pumping strategies throughout the system.  The speed of the collector pump P3 could be varied to maintain a set temperature gain across the collectors, eliminating cycling in low light conditions.  The speed of pumps P1 and P2 could be varied to maintain a constant pressure differential no matter which zone valves were open.  Often the added cost of variable speed pumps can be recovered quickly from savings in electrical costs.  Variable speed pumping also reduces piping wear by eliminating unnecessarily high flow rates.

Potential improvements aside, I like my solar heating system very much.  The collector area to storage volume is nearly ideal and the simplicity of the boiler interface will be hard to beat.  I still get a thrill from taking a solar heated shower and am amazed that free energy is there for the taking.  I hope you have found some useful information in this case study.  Future articles will cover some of the concepts mentioned in this article in more generality and depth. Specifically, I would like to write more about: payback, collector efficiency and control strategies.  

1   Tom Lane, Solar Hot Water Systems, N.p.: n.p., 2004, p.19.