Introduction
This section was developed to help share the reliability experiences of solar water heating professionals with other professionals and the public. This page was developed because reliability information is so so central to the success of solar water heating, yet so difficult to obtain.

Most of the information below was compiled from conversations with solar contracting and repair service people operating in the California Central Valley and Sierra Foothills. Comments or additional resources/links related to this subject would be appreciated very much.


Contents

General reliability issues

Repair, replacement & maintenance estimates

Resources

Links

General Reliability Issues (& solutions for dealing with them)

1) Freeze protection
2) Overheat protection
3) Poor water quality
4) Mixing valve and tempering valve failure
5) Lack of maintenance
6) Lack of a check valve
7) Pump failure
8) Loss of fluid
9) Suction head failure (drainback systems)
10) Siphon loop failure (drainback systems)
11) Entrained air
12) Glycol failure from stagnation (pressurized, closed loop systems)
13) Pressure or Pressure/Temperature (P-T) relief valve failure
14) Air vent failure
15) Lack of proper flow
16) Expansion tank failure
17) Expansion tank orientation
18) Expansion tank under-sized
19) Sensor wire failure
20) Temperature sensor failure
21) Limited hot water recovery capacity
22) Storage tank failure
23) Safety shutoff keeps turning system off
24) Faulty installation of system components
25) Improper dip tube installation
26) Failure of pipe insulation materials
27) Roof-related problems


1) Freeze protection
In areas where the temperature can get to freezing (and in most cases, even slightly above-freezing) temperatures, the method the system uses for freeze protection becomes critical. If water can freeze inside a pipe or tank it will split it open as the water expands while freezing.

Recirculation systems will freeze on pump failure and on power failure. Freeze problems are observed in multiple collector array systems with recirculation freeze protection, when the controller and pump appear functional. It may be due to scaling in some tubes, or flow imbalance in arrays, with "starved" risers at particular locations (see SOLDIS program). Recirculation freeze protection in large arrays is therefore considered questionable.

In drainback systems using small modules, sensor and controller failure can lead to frozen collectors. In severe climates, failure of the air vent valve to open in draindown systems causes draining to be too slow, and freezing can occur. Most important in a drainback system is to assure the collector and all supply/return piping is tilted to allow water to fully drain from the system when the pump is off; be careful with large collectors mounted horizontally, as the internal piping can sag over time and trap water.

Some contractors claim collectors made of EPDM can be routinely frozen, while others claim only limited cycles of freezing before a leak can develop. Polybutylene pipe is also sometimes characterized as "freeze-proof", however it may burst if routinely freeze-cycled, as evidenced from repairs of solar swimming pool systems with such piping that were not drained in winter.

Solution: Choose a system with an appropriate (proven) freeze-protection approach for your climate. Also, all exposed collector piping should be heavily insulated and wrapped with protective foil tape or other material where exposed to the outdoors. It is critical to periodically inspect and maintain the pipe insulation & protective materials to maintain freeze protection.

2) Overheat protection
A solar water heating system can reach very high temperatures when there is little hot water useage, if it is not operated for a day or longer, or if the storage tank is too small compared to the collector area. High temperatures also cause pumps and tanks to fail prematurely. High temperatures also accelerate problems associated with poor water quality (see item 3, below). When a system has experienced a lack of fluid flow, causing it to continue to heat up, the system has reached what is referred to as "stagnation." Solar water heating systems must be designed and operated properly to avoid frequent collector stagnation and premature failures.

Active solar water heating systems using a differential controller generally shut the pump off at a maximum preset temperature, usually 160-180 (degrees F). If the solar tank cools off just a little, the controller will switch on the pump and send very high temperature "slugs" of solution into the system, causing expansion stresses and general degradation. This leads to premature component failure (tanks, pumps) and piping leaks. System heat transfer fluids such as glycol will also fail prematurely if exposed to high temperatures over time.

Passive systems generally use a "temperature/pressure" relief valve to discharge water from the storage when it reaches 180 (degrees F); they aren't designed for "continuous duty," however, and will likely fail prematurely if the system is frequently in a condition of stagnation.

Solution: Purchase a system that is properly sized and installed for your climate and pattern of hot water use. If the system is used occasionally, such as in a vacation cabin, it should be covered or protected in some other way from overheating.

Some system types, such as drainback systems, handle occasional overheating just fine, and would be the system of choice for such situations. Passive systems using a "temperature/pressure" relief valve, as well as other systems which are prone to failure or high-maintenance, should be avoided in cases where hot water is not used on a regular basis unless steps are taken to prevent failure due to frequent overheating.

3) Poor water quality
Most solar water heating systems will begin to perform poorly if mineral deposits or debris collect in the system. Poor water quality leads to early pump and tank failure, ruins valves, and will cause collector failure in systems where cold supply water flows directly through the collector on the way to the storage tank (usually seen as freeze damage). Flushing the storage tank every year to remove sediment can help, but some systems are difficult to virtually impossible to take apart and clean. In addition to sedimentation, there are two basic types of water quality problems:

Corrosion
Corrosion results from acidic water and can be an acute problem in well-water sites, where water quality is not controlled. Solder flux, if not properly flushed out, may cause a closed loop fluid to become acidic, promoting corrosion.

Scaling
Scaling is a problem in areas with hard (high calcium carbonate content) water. Municipal water suppliers frequently add calcium carbonate (lime) to water to decrease corrosion potential, which unfortunately increases the potential for scaling. Certain system types are a problem in areas with hard water. Open loop recirculation systems or draindown systems are prone to collector scaling, which can plug the collector tubing causing lack of flow as well as increased potential for freezing from trapped water. Closed loop systems will have scaling at the heat exchanger, which will require regular maintenance to remove. High water concentrations of Manganese (Mn), a metal that rusts like Iron, and Iron (Fe), along with high temperatures, promote a reddish "bio-fouling" that can completly plug the header of a draindown system.

Under no-load conditions, passive systems will generally have uncontrolled summer temperatures which will cause the P-T valve to release hot water frequently. Besides leading to valve failure (P-T valves are not meant to cycle), scaling is promoted at these higher temperatures.

Solution: Always check the quality of the water supply you plan to use before selecting a solar water heating system. Make sure the inside of the system can be easily "de-limed," "de-scaled" and flushed of sediment in order to maintain system efficiency and service life. Clean the system as frequently as needed to keep sediment and deposits to a minimum. Also, consider a water conditioner or some other form of treatment if you have hard water. In general, larger, hotter-running systems have more scaling problems than smaller, cooler-running systems.

Water used for closed drainback loops must be neutralized if it is acidic, or it will corrode piping. With "shell-tube" type heat exchangers, the potable water side must be accessible for periodic de-scaling as this side is constantly exposed to fresh supplies of hard water and will therefore be most impacted.

4) Mixing valve and tempering valve failure
Solar water heating systems can generate temperatures in excess of 160 (degrees F) in summer or when the system is not used for a period of time. Mixing valves are an important safety device, as they mix cold water with the hot water supply in order to keep the hot water below scalding temperature. Recent national plumbing code changes require a mixing valve on all hot water lines serving showers, and other similar fixtures where danger of burns exist.

Certain "tempering valves," sometimes (incorrectly) referred to as "mixing valves," are a major maintenance problem, with mean lifetime as little as 3 or 4 years. Cold water delivery is generally the result of a tempering valve failure. Many times, the valve is "stuck" in the "cold open" condition due to scale build-up or debris in the line, and can be fixed by simply removing, cleaning, and exercising the spring mechanism. Eventually, the spring will corrode and break, or become so clotted with scale as to be inoperable. Repair is easy if the system design has allowed isolation of the valve; otherwise it is more difficult and expensive.

Solution: "Tempering valves" and certain other types of temperature control valves do not provide the proper "anti-scald" function and long-lasting quality of a mixing valve, and should not be used in place of one.

A mixing valve must be installed with a thermal trap, per OEM guidelines. Be sure the mixing valve is a properly certified "anti-scald" device and is properly installed. Although some "anti-scald" valve manufactures do not require a thermal trap, some solar contractors recommend a trap anyway, as it reduces the high "top of tank" temperature the valve would otherwise be exposed to, reducing the potential for scaling. In setting the valve position, direct measurement of water temperature (after running for several minutes) is recommended to catch miscalibration problems before they turn into a service "call-back."

When the mixing valve is between the solar system and the storage tank (as it is on many passive systems or systems with gas-auxiliary), it is very useful to place the solar outlet temperature probe downstream from the mixing valve, so that mixing valve failure is immediately evident. It is also useful to install valves to isolate the mixing valve for servicing.

When installing the valve, it is important to remove the regulation mechanism from the unit before the unit is sweated into place. The high soldering temperature (due to the requirement for non-lead solder) may be one cause of premature failure. When the mechanism cannot be removed, threaded units should be bought, with some provision, such as a wet cloth, for keeping the mechanism cool when the nearby joints are soldered.

Tempering valves can cause feed-through between hot/cold lines, especially in larger buildings with large hot water recirculation loops, as the pressures will generally not be equal on hot/cold side of the valve. This can be fixed through the use of a check valve or other solutions recommended by the manufacturer.

5) Lack of maintenance
Most solar water heating systems require some level of maintenance, if nothing more than flushing the solar storage tank every year (also recommended for typical gas & electric water heaters) in order to remove sediment and extend tank life.

Solution: Place a maintenance manual and/or maintenance sheet in a prominent location in the space where the solar storage or access door is located. A simple maintenance sheet encased in plastic is best for longevity, and is a sign of a consciencious installer.

Some systems can be maintained by a homeowner or maintenance person while others cannot. Know which type of system you plan to purchase, and provide the necessary maintenance to assure your investment is secure. Consider a maintenance contract with a reputable solar service person in your area, or call a solar repair person to maintain your system per manufacturers instructions.

6) Lack of a check valve
A check valve is sometimes recommended in the cold line leading to the mixing valve at the system outlet, to prevent hot water flowing back through the check valve and heating the cold water.

Solution: Install a check valve or a good thermal trap (or "U-tube") in the cold water line leading to the mixing valve.

7) Pump failure
How long a pump lasts depends on how cool it is allowed to operate. The hotter the fluid its constantly pumping, the sooner it will burn out. Aside from lack of use, other factors that increase system temperatures include oversized collectors (relative to the storage tank volume), and systems with selective surface solar collectors.

Solution: The pump should always be installed at the coolest place in the solar loop(s) to prolong its life (some manufacturers do not follow this rule). Also, the collector-to-storage ratio for the application must be properly selected.

In some cases, pumps have special installation needs. For example, Grundfos pumps must be installed with shaft horizontal as shown in the manufacturers installation instructions. Also, pumps that require 120V AC need electrical conduit over the wiring leading to the pump; use of 24VAC pumps does not require conduit.

Different solar contractors tend to have strong preferences towards one of the two primary pump manufacturers, Grundfos and TACO, although there is no clear consensus. One contractor has said the Grundfos pump has a carbon-steel shaft, which can wear, ride up the bearing sleeves, and "stick". The same contractor noted that he has replaced more Grundfos pumps than TACO, even though he personally installed four times more TACO than Grundfos. An advantage of TACO is that the core+propeller unit is an easily replaceable cartridge. Several other contractors felt Grundfos was more durable and the best pump choice.

8) Loss of fluid
Leaky valves or other connections can cause fluid loss. In systems with small drainback tanks (with a small vent to the atmosphere at the top of the tank), the combination of large collectors and a hot climate can cause high evaporation rates. With a dry tank or insufficient suction head (see "entrained air" below), the pump will burnout.

Solution: The water level in drainback systems should be checked regularly, especially at first in order to determine how frequently water should be added. The system should be inspected at least once a year to be sure no leaks exist at valves or other connections.

9) Suction head failure (drainback systems)
In drainback systems, careful attention should be paid to the head required on the suction side for proper starting. For example, the Grundfos series requires larger suction head for starting than the TACO pump. Some drainback manufacturers appear to place the pump immediately below the tank, and head may be insufficient.

Solution: The pump must be properly sized and selected by a trained professional.

10) Siphon loop failure (drainback systems)
Care should be taken in drainback systems with an inverted U-loop siphon from the tank to the pump. Any air entrapped in the U loop can cause failure to pump, and premature pump failure.

Solution: See "entrained air" below.

11) Entrained air
Pump failure can be caused by air within the solar piping loop becoming trapped in the impeller cavity, leading to loss of suction, free-wheeling and burnout; designs with the pump input/output ports installed horizontal are potentially subject to this problem.

Solution: There are several effective ways to purge air from the piping. Sources of information include the Hydronic Training materials, such as from the I=B=R institute.

12) Glycol failure from stagnation (pressurized, closed loop systems)
Glycol heat transfer fluids degrade and eventually fail more rapidly in systems which regularily encounter high stagnation tempertures. Systems having collectors coated with a selective surface further exacerbates the problem.

Solution: Lifetime of the glycol is increased when the collector is tilted to favor winter sun, versus summer sun. Weather also plays a role, as high summer outdoor temperatures with intense sunlight in the western U.S. (versus the cooler "cloudy summers" in the eastern U.S.) will tend to degrade fluid more quickly.

For larger systems, the homeowner can be encouraged to cover the collectors when going on an extended vacation to prevent stagnation damage of the glycol. The specific choice of heat transfer fluid can also help. For example, Dowfrost HD is four times more thermally stable than Dowfrost NT and is good for continuous service up to 325F, whereas Dowfront NT is good only up to 250F.

13) Pressure or Pressure/Temperature (P-T) relief valve failure
Many passive systems use a pressure or pressure/temperature relief valve at the collector/storage assembly outlet on the roof. These valves are set to expel water from the system whenever the temperature and/or pressure exceeds a fixed valve. If a pressure relief of 75-80 PSI is used, the "service pressure", which can be as high as 100 PSI or more, will cause the water in the system to discharge. Another problem can arise when the pressure reducer is on the hot water system inlet only, the cold (at potentially higher pressure than the pressure relief) will often bleed through the common "one-armed" mixing faucet fixtures, pressurize the hot water side, and set off the pressure relief in short, repeated bursts.

With P-T valves, a system that has little to no hot water use for a period of time may tend to overheat, causing the temperature relief to expel water frequently. Because P-T valves are not intended for frequent use, the temperature at which they turn on is reduced with useage, exacerbating the problem.

Solution: A pressure regulator must be used to reduce pressure whenever the service pressure is higher than the system relief pressure. The regulator should be installed and set to the pressure recommended by the manufacturer.

Code officials sometimes require P-T relief on the collector, even though the valves adequately cover safety concerns when located in the loop inside near the tanks. When they are installed on the collector outlet, this leads to unnecessary fluid dumping under high temperature conditions. One solution satisfying the building inspector is to place the required valve on the bottom header. And finally, installers should always be careful to place the outlet of the pressure or P-T relief valve in a safe place to avoid the potential of a discharge of hot water burning a person, a pet, or the garden!

14) Air vent failure
In general, failures of air vent valves is seen as a real maintenance problem by most contractors. On pressurized closed-loop systems, the valve is intended only for air venting during and immediately after filling. Several solar contractors maintain air vent valves are unnecessary, if air is properly purged from the system initially. One solar contractor believes that air vent valves have close to a 100% failure rate within 5 years

When air vents fail the system looses fluid from the loop, fails to operate, and damage to the system or the mechanical space can result. For glycol systems, solar fluid may boil under "no load" circumstances. If present, an air vent valve may release fluid vapor, and/or the pressure relief (set at low value of 50 psi on one system type) may exhaust fluid, leading to low fluid levels and potential vapor-lock.

Solution: Use a charging pump with container to evacuate the air from the system and back into the container using the pump to pressurize the system. Also refer to system installation manual for other recommended solutions.

15) Lack of proper flow
A proper flow rate can be crucial to system efficiency, yet many manufacturers do not provide flow meters with their systems as a cost saving measure. Given the possibly large variances in installed piping lengths and static head (pump lift height) for different buildings, the potential for debris (e.g., copper plugs from header holes, blobs of solder, debris in the water, etc.) to block passageways; it's difficult to verify proper performance and fine-tune the system without a flow meter.

Solution: If a flow-meter does not come with your system, you should ask the manufacturer and/or installer to include one. A plastic ball-float flow meter installed in the line can be an inexpensive and reliable solution. Also, a strainer in the loop can prevent some of the potential flow blocking from debris. One negative aspect to flow meters is that some will eventually leak and need to be replaced. Be sure the owners manual for the system you purchase describes a "homeowner test" for assuring proper solar system function.

16) Expansion tank failure
The lifetime of an expansion tank is generally a function of the tank material and bladder design. Inexpensive tank or bladder designs will fail much sooner than those made of higher quality materials.

Solution: A quality stainless steel tank will last longer than a steel tank. For bladder lifetime and flexibility, some solar contractors maintain there's an advantage to pure nitrogen charge of the tank (rather than air).

17) Expansion tank orientation
There is some disagreement on the "proper orientation" of expansion tanks as to whether they should be installed upright, horizontal, or up-side down. One solar contractor points out that in glycol loops it is impossible to purge the air unless the tank is in an upside-down configuration; this is made clear in general hydraulic training.

Solution: Follow installation instructions for the expansion tank from the tank manufacturer unless the solar manufacturer and/or installer presents a convincing argument for an alternative.

18) Expansion tank under-sized
When expansion tanks are undersized they do not have sufficient capacity for the expansion of fluid as it heats up within long piping runs. Excessive system expansion can cause pressure-sensitive components to fail, resulting in fluid loss, system failure, and damage to system components and the mechanical space.

Solution: Manufacturers of expansion tanks have tables relating the volume of expansion tank to the volume of fluid in the system. These tables should be used by the installer to assure the expansion tank is the correct size given the fluid capacity of the system and piping.

19 ) Sensor wire failure
The wiring that comes out of the body of a freeze snap switch can be easily stressed and broken if handled too roughly during installation, causing the switch to fail. Steel clamps on sensor wire can short the sensor loop, leading to 24 hour pump operation. Occasionally, electrical interference from a nearby radio transmitter can cause rapid pump cycling.

Solution: When mounting freeze snap-switch with wires directly entering the casing, care should be taken to relieve stress on and prevent excess motion of the wires, which can lead to stress breakage of the lead wires at the sensor body. Only stranded, UV-protected sensor wire should be used. Telephone wire or bell wire will fail. Use plastic zip-ties or other (softer than steel) clamps to hold the wiring in place. Electrical interference can be solved by using shielded wire for sensor lines. The shield should be grounded at only one end, probably at the controller. Also, be sure not to attach sensor wire to collector piping, as the heat will melt the insulation.

20) Temperature sensor failure
Some manufacturers temperature sensors are prone to early failure.

Solution: Check with an experienced solar contractor to find out which manufacturers are currently making reliable sensors.

21) Limited hot water recovery capacity
Consumers must clearly understand the limitations and appropriate use of the Rheem (wraparound) tank as well as most single tank (electric backup) systems. On any single-tank systems with small electrically-heated volumes, the homeowners will experience reduced quantities of hot water during cloudy periods.

Solution: Sales and installation personnel must not distort the potential for running out of hot water. Misunderstandings cause frequent and costly call-backs and detract from the value of solar water heating in the mind of consumers. Other solutions include the installation of a larger electric element (5 and 6kW are available; be sure code regulations for wire size are met), turning up the thermostat (e.g., 150F), and the installation of low-flow fixtures (which also conserve water). Be advised, however, that turning up the thermostat will lower the solar fraction due to the extra heating of the tank.

Some have suggested rotating the existing element downward to affect a larger volume of water, however this raises an unresolved question as to stress weakening and element failure, although some solar contractors claim no elements have failed as a result of this practice.

22) Storage tank failure
Solar contractors have discovered that fiberglass (glass-lined steel) storage tanks have resins in the material which leach out at higher temperatures, raise pH, and cause leaks in copper piping. Also, fiberglass tank lifetime is shortened by exposure to high temperatures which are commonly up to 160 (degrees F) in solar applications; water heaters are generally set to just 130 to 140 (degrees F).

Early Rheem "wrap around heat exchanger" type storage tank designs were a problem for solar contractors; 1 in 5 to 1 in 10 tanks suffered leaks on the collector loop wrap-around plumbing. Later tanks seem much better, as Rheem apparently identified the problem as a poor solder joint at the end of the wrap around section (installers beware). Also, the Rheem tanks plastic drain fitting frequently leaks, especially at higher line pressures (which can be 100 psi in the some areas). And finally, there is a thermometer well in the side the Rheem tank with a threaded plug inserted at the factory. Removal of the plug to insert an immersion type thermometer creates additional labor and hardware cost, and has been found to be located too high on the side of the tank to be an effective control location.

Welded stainless steel tanks, often characterized as the highest quality solar storage tank material, can fail just as soon or sooner than fiberglass tanks, depending on the shape of tank, tank wall thickness, and quality of weld (metal has a high coefficient of thermal expansion which, depending on tank shape & wall thickness, can break welded joints). High temperatures also accentuate problems caused by thermal expansion.

Some have reported EPDM lined tanks containing copper heat exchangers have have devoped premature heat exchanger failure.

Solution: Weigh out the cost of fiberglass storage tank replacement and, potentially, repair of leaks in copper piping against the cost of a tank made of a longer-lasting, more heat resistant material. An economic analysis, such as a life-cycle cost calculation, will put the time-value of money relationship of these costs in proper perspective. The life of a fiberglass tank can be extended significantly if the temperatures are kept lower and the anode rods are checked regularily and replaced when necessary; this is true for water heaters in general. Also, some solar contractors believe the storage tank should have an expansion tank if a pressure reducer is used on the supply, as thermal expansion can cause small fractures, leading to rust and leaks.

For placing sensors on Rheem tanks, metal strapping on the drain nipple is much quicker, causes less reliability problems, and is sufficiently accurate. In any case, the control sensor should be located at the bottom of the tank so as to sense the coolest temperature.

There are reports of stainless steel tanks that have lasted the test of time quite well; your best bet with a stainless tank is to select one with successful installation records of 10 to 15 years (or more) without problems.

Use of EPDM lined tanks should be approached carefully. Be sure the manufacturer has sufficiently addressed long term reliability issues, or that the maintenance and replacement costs associated with liner replacement are not excessive.

23) Safety shutoff keeps turning system off
The use of a sensor in the bottom of the tank for high-temperature cutoff is a potential problem. When this is set too near 180 (degrees F), there is a potential for tripping the "high temperature safety shutoff" (apparently also 180) on tanks with an auxiliary element. The problem also exists potentially with two-tank systems after the system has fully heated and a large draw is very quickly taken (e.g., summer vacation return).

Solution: Some solar contractors have suggested that there should be some way for the safety reset to trip only when power is applied, in order to distinguish between a malfunction of electrical elements and an overheated hot solar tank (not its job).

24) Faulty installation of system components
Some manufacturers do not provide complete "modules" with their system, requiring the installer to piece together the heat exchanger, pump, controls, and other components to build the system up. These types of systems risk component failures should the installer put them together incorrectly.

Solution: Manufacturers should "modularize" the tank, heat exchanger, pumping and controls, rather than asking the installer to do so. If the manufacturer delivers pieces, they are best pre-assembled in the shop before the system is assembled in the field. Avoid systems that have been "modularized" to the extent that individual components, that may require service or replacement at some point in time, are difficult to impossible to service; these systems will have unacceptably high service costs.

25) Improper dip tube installation
Dip tubes are too frequently "unattached" or missing; also, it can become unattached when thermosiphon nipples are installed. When the dip tube is not properly present, the user may experience cool water, even when the tank is fully charged.

Solution: The dip tube should be checked before field installation (where solution is a major chore); removing the element allows good observation of the tube.

26) Failure of pipe insulation materials
Because operating temperatures differ significantly with solar system design it is important to select an insulation material that can withstand the temperatures involved. Failure to do so can result in melted insulation. Insulation exposed to weather will degrade quickly if not protected from the sun's ultra-violet (UV) rays, water, and the claws of cats or other animals. Improperly installed insulation will shrink, leaving gaps through which the weather and insects can enter.

Solution: Pipe insulation protection is best done using aluminum foil tape with special adhesive. It is less labor-intensive than two coats of UV-inhibiting paint and lasts longer. Job sites show the tape in good shape after ten years of exposure. Some solar contractors believe painting is altogether unsatisfactory. Those who stand by the use of paint emphasize the need for two brushed-on (NOT sprayed) coats, and the use of semi-gloss, light colored paint.

When installing pipe insulation, remember it will tend to shrink slightly. To avoid this problem, the installer should longitudinally compress the insulation to prevent separation at the joints in the insulation sections, upon shrinking.

27) Roof related problems

Roof leaks
Roof leaks are not uncommon on wooden shake shingles, especially older, less flexible wood.

Solution: One solar contractor uses flat metal gutter flashing around each penetration, about 1 square foot, to stop leaks through cracks; he says caulking applied after the wood screw is set doesn't work. Instead, he fills the pilot hole with caulk initially, and then drives the wood screws in. Another solar contractor says that when the wood screw is properly set into a joist rafter and caulked before screwing, leaks are rare. In fact, he said, use of flashing can cause tears in the tar paper, which is what blocks water in the first place, not the shingles.

Generally speaking, avoid the use of wood "sleepers," as wood tends to weather poorly. In a few rare instances where a deciduous tree has dropped a lot of leaves behind the collector and the collector was mounted too close to the roof, fires have started in the leaves and/or wood sleepers when the system was left unused for a period of time (and overheated). Most solar contractors recommend some form of metal mounting hardware for durability and longevity.

Re-roofing
A number of systems are simply removed at re-roofing, as the roofing contractors do not feel comfortable re-installing the systems.

Solution: Collector or collector/storage systems installed into the roof like a skylight is one solution around removal, although guarding against leakage given the expansion/contraction of the unit is very difficult in practice. Another solution is to use mounting hardware that allows the collector to be simply "hinged up" on one side, after piping has been cut; after re-roofing, it's then hinged back down, bolted, and the piping is re-connected. Ask your system manufacturer and installer about other suggestions they may have for installation that would not require complete removal of the collector for re-roofing.

REPAIR, REPLACEMENT AND
MAINTENANCE ESTIMATES

Introduction
The information in the following tables is taken from a survey of California Central Valley & Sierra Foothill Solar Contractors. Due to the limited number of survey respondents upon which the data in these tables is based, the information is of limited value to those who require a statistically valid sample size or for those wishing to make general conclusions regarding the reliability of a specific manufacturers component or system.

Table 1 - Component Mean Lifetime "BEST CASE" Conditions
Well-maintained, good water quality, and properly sized systems (lower operating temperatures).

Table 2 - Component Mean Lifetime "WORST CASE" Conditions
Poor water quality and aggressively sized systems (with higher operating temperatures).

Table 3a - Self-reported Cumulative Service Call Numbers
Contractors #1-4

Table 3b - Self-reported Cumulative Service Call Numbers
Contractors #5-9

Component	Low(2)	High(3)
Collector
Glass cover	30	60
Polycarbonate cover	5	20
Plastic films (Tedlar)	5	20
Copper absorber	20	60
EPDM absorber	5	20
Glycol fluid	5	10
Gaskets	?	?
Tanks
Glass-lined	8	25
Polypropylene (unpress.)	20	40
Pumps	5	20
Controller
Current models	10	30
Sensors	10	20
Loop regulation
Mixing valve, no trap	3	7
Mixing valve, trapped	5	30
Check valves	10	40
Vent valve	3	8
Vacuum relief	3	10
Draindown valve	3	9
Expansion tank	5	20
Pressure relief valve	10	25
Pipe insulation
painted	2	8
aluminum tape	8	10

Notes
1) Mean lifetime: The time for 50% of the population of operating units to fail.
2) Low: lowest estimate provided by contractors surveyed.
3) High: highest estimate provided by contractors surveyed.

Table 2
Component Mean Lifetime Estimates (1)
Survey of California Central Valley & Sierra Foothill Solar Contractors

"WORST CASE" Conditions
Poor water quality and aggressively sized
systems (with higher operating temperatures).

Component	Low(2)	High(3)
Collector
Glass cover	30	60
Polycarbonate cover	5	20
Plastic films (Tedlar)	5	20
Copper absorber	10	30
EPDM absorber	5	20
Glycol fluid	3	6
Gaskets	?	?
Tanks
Glass-lined	5	20
Polypropylene (unpress.)	10	20
Pumps	3	10
Controller
Current models	10	30
Sensors	10	20
Loop regulation
Mixing valve, no trap	2	5
Mixing valve, trapped	5	10
Check valves	5	10
Vent valve	2	6
Vacuum relief	2	6
Draindown valve	2	6
Expansion tank	2	6
Pressure relief valve	4	12
Pipe insulation
painted	2	8
aluminum tape	8	10

Notes
1) Mean lifetime: The time for 50% of the population of operating units to fail.
2) Low: lowest estimate provided by contractors.
3) High: highest estimate provided by contractors.

Contractor #	1	2	3	4
#Systems(2)/AvgYrs(3)	125/1.5	2000/11	500/4	350/2
System type	Glycol	Draindown	Film ICS	Drainback
Cause for Service Call:
Collector glazing	0	?	3 (5)	0
Absorber	0	80 (9)	na	0
Tank	1	25	na	2
Pumps	0	350	na	2 (8)
Controller	0	200	na	1
Loop regulation
Mixing valve, no trap	3	400	?	2
Mixing valve, trapped	na	na	na	na
Check valves	0	na	na	0
P-T relief valve	? (7)	?	na	1
Vent valve	na (4)	some	?	na
Vacuum relief	na	many	?	na
Draindown valve	na	850	na	na
Expansion tank	0	na	na	0
Temperature probe leak	5	na	na	0
Piping	0	20	2 (6)	0
Homeowner education
Single-tank recovery	3	0	0	?

Notes
1) Total number of service calls reported for the entire base of systems.
2) The total number of systems in a given, well-defined base of systems for which the contractor is solely responsible.
3) The average age of the systems in the base; for SMUD program systems, the number is less than three, and roughly indicates the time history of activity.
4) Stopped installing them on systems, used brass "T" w/plug for filling.
5) About 25 damaged glazing films replaced from "trees, rocks..."; and about 1% seam failure on the early stainless steel model.
6) Pipes burst under unusually low temperature event and no use by occupants.
7) Had some trouble on other systems with P-T on collectors when required. Need to be placed on inlet header, not outlet.
8) Early design of drainback tank caused some leaking, which led to dry tank and 2 pump failures.
9) Extreme freeze damage, failed vacuum relief and/or draindown valve.

Contractor #	5	6	7	8	9
#Systems(2)/AvgYrs(3)	350/1.5	100/5	500/4	200/1.5	250/2
System type	Glycol	Drainback	Thermosiphon	Glycol	ICS (12)
Cause for Service Call:
Collector glazing	0	?	1	0	few (13)
Absorber	0	0	0	0	na
Tank	20 (5)	few (9)	0?	40 (10)	na
Pumps	2	0?	na	0	25
Controller	0	?	na	0	12 (14)
Loop regulation
Mixing valve, no trap	45 (6)	0 (11)	?	2	60
Mixing valve, trapped	na	0 (8)	na	na	0
Check valves	0	some	?	0	?
P-T relief valve	0	?	na	1
Vent valve	na	?	?	?	na
Vacuum relief	na	?	na	?	na
Drainback valve	na	na	na	na	na
Expansion tank	0	0	na	0	0
Temperature probe leak	5	na	?	0	0
Flow meter break/leak	na	0	na	na	na
Lack of high limit	130 (7)
Piping	many (4)	?	0	0
Homeowner education
One-tank recovery	?	Many?	0	?	?

Notes
1) Total number of service calls reported for the entire base of systems.
2) The total number of systems in a given, well-defined base of systems for which the contractor is solely responsible.
3) The average age of the systems in the base.
4) Experienced leaks with charge/fill hose bibs, leading to loss of loop pressure. Homeowner would call if P<10psi. Re-tightening of these bibs was found required to prevent leaking. A higher quality boiler drain valve resolved the problem.
5) Many failures (about 25/350), early on especially, with the Rheem "wraparound" tank. The main problem was the heat exchanger leaking at the bottom or top fittings. Although most leaks were found in the shop, about 20 failed in the field during the first week of operation.
6) Due to a calibration problem using a tempering valve, experienced by other installers also. Some problems with "too much cold", stuck in cold on position. Problem was resolved by using a mixing valve.
7) These systems were mistakingly installed without a high limit setting.
8) Also uses the mixing valve, not the tempering valve.
9) Polypropylene tank. Had some early problems with the weld, but material seems to last indefinitely.
10) High collector loop leak rate on Rheem tank, early on. Most rejected in the shop.
11) Many repairs reported, but not on the installed based of this type of system which has a thermal trap between tank outlet and mixing valve.
12) ICS unit with hybrid pump added on later to help in overheating problems.
13) Several tubes were observed to be broken, probably vandalism.
14) Thermal snap switch, set to 160F, replacements.

Reliability data resources
1) Dr. Byard Wood, Reno State University, Nevada. Completing a solar reliability study (as of October 1997).

2) "1994 DOE Reliability Task Report (Draft), " Jay Burch, National Renewable Energy Laboratory; 1st Draft 2/16/94; 1st Revision 6/30/94; 2nd Revision 9/23/94.

3) "Solar Water Heating - Well-Proven Technology Pays Off in Several Situations," FEMP (Federal Energy Management Program) Federal Technology Alert - A series of technology guides prepared by the New Technology Demonstration Program, U.S. Department Of Energy's Strategic Environmental research and Development Program; May 1996. For more information contact FEMP's help line at 1-800-DOE-EREC or use the link to FEMP, below).

4) "Field Performance of Solar Water Heating Systems - Residential Systems (Draft)," Colleen Kettles and Tim Merrigan, Florida Solar Energy Center; July 1994.

5) "Software to Predict Scaling in Solar Domestic Hot Water Systems," Dr. Gary Vliet et.al., University of Texas At Austin, Department of Mechanical Engineering, prepared for The National Renewable Energy Lab under Contract #DE-FG36-94 G010034; November 1995.

6) "A Method for Including Operation and Maintenance Costs in the Economic Analysis of Active Solar Energy Systems," Walter D. Short, prepared for the U.S. Department Of Energy under task #3002.10 by the Solar Energy Research Institute (now called the National Renewable Energy Laboratory, NREL); August 1986.

7) "Solar System Reliability Survey Results," Gary Curtis et.al., Oregon Department Of Energy; Undated (appears to be mid - 1980's).

8) "Final Reliability and Materials Design Guidelines for Solar Domestic Hot Water Systems," ANL/SDP-11, Argonne National Laboratory, September 1981.

9) "Summary of Performance Problems of 100 Residential Solar Water Heaters Installed by New England Electric Co. Subsidiaries in 1976 and 1977, " Robert O. Smith and Associates; October 1977.

10) "An Owner's Manual For Your Solar Hot Water System; Type III: Drainback," Steve Shewmake, 10607 Bragg Avenue, Grass Valley, CA 95945, or email Steve @ solar@jps.net. This book, like the others he's written covering other system types, addresses how the system works, maintenance, trouble-shooting, and how to optimize your systems performance.

11) "A Solar Repairmen's Notes on Reliability," Steve Shewmake, 10607 Bragg Avenue, Grass Valley, CA 95945, or email Steve @ solar@jps.net. This document summarizes the solar water heating lessons learned, solutions providied, and repair/maintenance costs from a solar installer/repairmen's five year history.

12) "Wisconsin's 'Orphan' Solar Program," Jeff DeLaune, Chip Bircher and Richard Lane, Home Energy Magazine, May/June 1995. One Utility found that rehabilitating old solar water heaters amounted to a cost-effective energy conservation program.

Reliability Data Links

Solar Hot Water Systems: Lessons Learned 1977 to Today -
The Solar Industry's Water Heater Bible
"Finally, a definitive how-to book for installing and maintaining high-performance and low-maintenance solar hot water systems -- written by one of the leaders in solar contracting today!"

Active Solar Heating System Reliability
This is the first web-based link I've been able to find to solar reliability. The page offers an explanation of common problems as well as a nice collection of references.