08-04-2016, 08:18 PM
First, tank-less water heaters have been somewhat over-hyped as though they have no flaws and are the answer to every conceivable problem. They have their place but they also have drawbacks. They must modulate the gas flow to match fluctuations in the water flow rate if a stable temperature is to be maintained. It is a fact of life that that is an imperfect process, more so on the less expensive models. They all seem to have the feature that if the flow and/or pressure drop too low the gas gets cut off, probably both to keep from producing slugs of scalding water and to keep from burning up the heater. I had no problems until I had a lot of company during the spring when the algae was blooming. The resultant loading up of my filter cut the flow rate of water and suddenly I was getting complaints that the water was too hot until it suddenly got cold. The solution was to clean my filter to get more flow. Actually in my case I have made several upgrades that go beyond solving the immediate problem. Some may recall that I have the Sawyer PointONE filter. It filters out anything larger than 0.1 microns. It was hugely expensive but will pay for itself over the long haul as it requires no energy (apart from pumping costs to force water through under pressure) and it is supposed to last 10 years. No annual cost for a UV bulb either. Naturally the stuff it filters out accumulates on the inlet side of the filter and must be removed periodically by backflushing. I had not done this for months and frankly I wonder that I did not have problems sooner.
A typical shower head flows around 1.5 gpm to 2.5 gpm. For the sake of argument lets say my shower head flow rate was around 2 gpm. My single Shurflo 4008-101-E65 pump claims to put out 3 gpm. It also claims to put out 55 psi. My experience with the pump confirms both figures, just not at the same time. As I troubleshot the problem I saw that while the pump was pushing water into the inlet of the filter at close to 55 psi, the pressure at the shower head was barely more than 20 psi. That's more than 30 psi drop across the filter and with only 20 psi at the shower head I was probably getting less than 1.5 gpm flow. My next step was to double up on the pumps, running 2 pumps in parallel. Come to think of it the figures above may have been with the second pump already in place. The pumps acted weird together so I controlled both pumps off of the same switch. They behaved better then but made a lot of noise as the pump pulses interacted. I kept turning up the pressure controls on the pumps until they finally would not shut off. Couldn't figure that one out until I went into the bathroom and discovered that the pressure relief valve on the water heater was relieving pressure all over the floor.
That whopping pressure drop across the filter was really kicking my okole. The filter manufacturer claimed that gravity flow of only 5 ft of head would result in nearly 20 gallons per hour, much less than what was required for a shower but if I piped that to a clean-water tank I could pump from that tank without the restriction of the filter at all. The clean-water tank would re-fill overnight using only gravity and hence no energy. This would be a vast improvement over a pump struggling to put out 55 psi but only getting 20 psi results at the shower head.
Long story short that's what I did and it seems to work pretty well. There were setbacks such as the hours I spent staring at the pathetic trickle of water coming out of the filter despite 6' of water in the tank, repeatedly flushing the filter to no avail. Only when I attached a hose to the filter outlet and saw that the trickle not only slowed but reversed as I lifted the end only half way up the tank did the gurgling noise coming from the pipe leading from the tank make sense. I had lost prime and the estimated 1 gph flow rate was due to having only inches, not feet of head to push the water through. I fixed that problem and now I am getting close to 20 gph when the catchment tank is full and the clean-water tank is low. As they equalize the flow rate decreases but as that means there is more water in the clean-water tank that is OK.
I should also say that I bought a new Dankoff model 2910-48 rotary vane pump that operates directly off my 48 volt battery bank. While the old Shurflo pump experienced a huge decrease in flow rate as the pressure built up the Dankoff 2910 just slows down slightly, never straying far from 4 gpm. Even sucking water through a 5 micron filter and feeding water to the shower head at a true 50 psi the new pump more than keeps up with the shower and shuts off periodically.
A typical shower head flows around 1.5 gpm to 2.5 gpm. For the sake of argument lets say my shower head flow rate was around 2 gpm. My single Shurflo 4008-101-E65 pump claims to put out 3 gpm. It also claims to put out 55 psi. My experience with the pump confirms both figures, just not at the same time. As I troubleshot the problem I saw that while the pump was pushing water into the inlet of the filter at close to 55 psi, the pressure at the shower head was barely more than 20 psi. That's more than 30 psi drop across the filter and with only 20 psi at the shower head I was probably getting less than 1.5 gpm flow. My next step was to double up on the pumps, running 2 pumps in parallel. Come to think of it the figures above may have been with the second pump already in place. The pumps acted weird together so I controlled both pumps off of the same switch. They behaved better then but made a lot of noise as the pump pulses interacted. I kept turning up the pressure controls on the pumps until they finally would not shut off. Couldn't figure that one out until I went into the bathroom and discovered that the pressure relief valve on the water heater was relieving pressure all over the floor.
That whopping pressure drop across the filter was really kicking my okole. The filter manufacturer claimed that gravity flow of only 5 ft of head would result in nearly 20 gallons per hour, much less than what was required for a shower but if I piped that to a clean-water tank I could pump from that tank without the restriction of the filter at all. The clean-water tank would re-fill overnight using only gravity and hence no energy. This would be a vast improvement over a pump struggling to put out 55 psi but only getting 20 psi results at the shower head.
Long story short that's what I did and it seems to work pretty well. There were setbacks such as the hours I spent staring at the pathetic trickle of water coming out of the filter despite 6' of water in the tank, repeatedly flushing the filter to no avail. Only when I attached a hose to the filter outlet and saw that the trickle not only slowed but reversed as I lifted the end only half way up the tank did the gurgling noise coming from the pipe leading from the tank make sense. I had lost prime and the estimated 1 gph flow rate was due to having only inches, not feet of head to push the water through. I fixed that problem and now I am getting close to 20 gph when the catchment tank is full and the clean-water tank is low. As they equalize the flow rate decreases but as that means there is more water in the clean-water tank that is OK.
I should also say that I bought a new Dankoff model 2910-48 rotary vane pump that operates directly off my 48 volt battery bank. While the old Shurflo pump experienced a huge decrease in flow rate as the pressure built up the Dankoff 2910 just slows down slightly, never straying far from 4 gpm. Even sucking water through a 5 micron filter and feeding water to the shower head at a true 50 psi the new pump more than keeps up with the shower and shuts off periodically.