Corrosion Chemistry in Solar Thermal Systems
Inside this Article
Many plumbers and solar thermal system installers have a limited background in corrosion avoidance. They rely on equipment manufacturers to engineer systems with long operational lives, and most do an excellent job. However, even a carefully engineered system can be installed incorrectly, and local water quality issues can result in the corrosion of components and premature system failure. Storage tanks made of steel and some types of stainless steel can be weak links in the system as they may be subject to corrosion failure.
Steel tanks utilize a sacrificial anode to slow the impacts of corrosion. Anodic and cathodic materials are relative to each other based on the electromotive force series, presented in Table 1. Aluminum, magnesium and zinc are all sacrificial and anodic compared to iron (the main component in steel). Sacrificial anodes, typically of aluminum, magnesium or zinc, are incorporated into virtually all glass-lined steel pressure storage tanks manufactured in the US. Once the rods are spent, the iron becomes sacrificial and anodic compared to any copper in the system and therefore begins to corrode rapidly.
The rate at which corrosion takes place is a function of the concentration of ionic species in the water, such as calcium, magnesium and sodium. These dissolved elements increase the conductivity of the water in the tank. Softening the water does not decrease its conductivity, but simply replaces the calcium and magnesium with sodium.
Iron from the tank’s steel wall will give up electrons and go into solution if it is electrically coupled to the copper piping in the house, which is typically grounded to earth. Most hot water tanks are dielectrically isolated from the house plumbing, precluding copper-based galvanic attack of the iron. The sacrificial anode is electrically connected to the iron in the hot water tank wall. Hence the aluminum or zinc gives up electrons more easily than iron, which gives the iron a negative charge so it cannot give up an electron and go into solution. This protects the iron in the steel from dissolving.
When I was on staff at the National Renewable Energy Laboratory, we would isolate the sacrificial anode with a PVC coupling. By hooking up an electrometer between the tank and the sacrificial anode, we could measure the coulombs transferred and calculate the dissolution rates of the anode aluminum or zinc. We could then predict the life of the anode as a function of tank temperature by connecting to copper pipe and a nonisolated recirculation pump. We concluded that as long as the tank stays electrically isolated, the sacrificial anodes can last up to 40 years. Electrical isolation is the key.
During service calls, I inspect the anode by inserting a Milwaukee M-Spector TV camera with a light on a waterproof probe into the hot water outlet. If the anode appears to be toward the end of its life, I install a new anode, and the tank is good for another 7 years or more. If the anode is so corroded away that the center steel rod is showing, I know the tank is compromised and needs to be replaced.
While stainless steel has a reputation for durability, this is not always the case. There are numerous and widespread instances of stainless tanks used in thermal systems experiencing premature failure. A common occurrence in these systems is failure at a tank’s welds.
Stainless steel is manufactured in hundreds of types. Numbers 304 and 316 are two common types used in tanks and heat exchangers. When 304 stainless, which is about 4% carbon, is welded, the carbon precipitates in the heat-affected zone in the form of chrome carbide. As a consequence, the mild steel in the heat zone corrodes through in a few years, even though the tank body will last for 40 years.
There are two solutions to this problem. The best one is for manufacturers to use 304L, or low carbon stainless, in tank manufacture. With only 0.5% carbon, chrome carbide does not precipitate, and the whole tank will resist corrosion for a long time. The second solution is for manufacturers to heat treat the entire tank after welding to about 1,400ºF to get the chrome back into solution, then quench the tank quickly in water so no chrome carbide can precipitate.
Number 316 austenitic stainless steel is also subject to chrome carbide precipitation in the weld–heat-affected zone. The problem can be solved by using 316L—also known as marine stainless, a favorite of sailors. A simple spark spectrograph can determine the chrome, nickel and carbon content in an unknown stainless steel. All manufacturers should utilize one to check incoming materials.
Since many expansion tanks are installed in a copper system and are not electrically isolated, they can also fail prematurely. Of the failed units I have observed, in all cases they corroded through on the band around the tank where the rubber bladder was sealed in place. It appeared that the bladder attachment process allowed cracks in the polymer protection coating, exposing the tank to corrosion. The expansion and contraction of the bladder seemed to push the iron oxide that was protecting the surface from further corrosion into the tank, increasing corrosion and adding iron oxide flakes into the system that in turn foul pumps and valves. In seal-less magnetically coupled pumps, these flakes accumulate on the rotor, slowing it down or stopping it, and result in poor performance or failed systems.
—Barry L. Butler, PhD / Butler Sun Solutions / Solana Beach, CA / butlersunsolutions.com