Solar Panels and Vacuum Breakers

[mas985]

To avoid a more technical discussion in this post, I moved the discussion to here. These are my comments to tphaggerty.

Sorry, you are just wrong! The breaker is a check valve. It is ALWAYS closed unless pulled open by a vacuum. The pump doesn't "close" the breaker when it turns on, it is already closed. There is NO WAY that the water "above" the breaker on the source side is "pulled up to the return line" (unless you have a hard closed source line). It just isn't going to happen that way. Think of the breaker as a one-way hole in your pipe, it doesn't matter WHERE the hole is, if the water is under vacuum, air will flow into the pipe. The only difference that the location of the breaker makes is how easily it opens. The lower it is, the less vacuum pressure it is under when the pump stops, the "harder" it is to open. That is why moving the breaker DOWN the pipe can help if you are getting air in the system when it is running. I agree that it makes sense to locate the breaker high, but if doesn't have to be for the system to work correctly.

I'm not sure where you get your information but vacuum breakers that I have seen are not check valves and are not always closed. The old style use a ping pong ball which floats to close the air vent when water reaches it and have been used for years. There are newer ones which work in a similar way but they all close with positive pressure on the inside to the pipe. This only occurs when water reaches the valve. I have been at the top of my roof when the solar is turned and I can assure that air escapes the valve until water reaches it. I have also taken apart an old valve to see how it works.

Secondly, you are completely ignoring how hydraulics work. I have a more detailed discussion below.

As for the air not being able to displace the water in the "very small pipes in the panel", they aren't that small, believe me. Most panels have panels have internal piping that is about 1/8 to 1/4 in diameter. About twice as big as a normal straw. Air will have no trouble rising up through the water in pipes that size to relieve the pressure.

So have you ever played the soda straw trick? Put a soda straw in a tall glass of water, place your thumb over the end and then pull out the straw. The water should stay in the straw unless you let air into the TOP of the straw. Just because there is air in the bottom of the straw does not mean that the water can flow out.

Go back and look at the powermat site videos. When the pump stops, the water drains down from the top of the system, air is allowed into the system and allows both sides to drain. And your explaination of why the water flows up because the pressure is lower at the top of the system makes no sense when you are allowing AIR into the system. Put a tennis ball at the bottom of the pipe, sure, it will be sucked up as the water drains. But if you OPEN the bottom of the pipe (the equivalent of opening the breaker), do you really think the water is still going to flow up the pipe?

This is the animation that I was talking about. Note that the panels drain from bottom to top.

So a little refresher on hydraulics seems in order. Static head is the most important thing here. Let's start with a thought experiment to illustrate what is going on.

We have two 23' pipes, one is capped at the top one is capped at the bottom. Both are filled with water but the one capped at the top is also in a bucket of water. The one capped at the bottom wil have a PSI pressure at the top of the pipe of 0 PSI since the top is open. At the bottom, the pressure will be +10 PSI.

For the second pipe that is capped at the top, the pressure at the bottom is now 0 PSI since it is effectively at atmospheric pressure of the bucket of water. However, the pressure at the top of the pipe is -10 PSI. There is still a 10 PSI difference but because the reference has change so has the absolute pressures. By the way, a 2" PVC pipe behaves in the same way as a straw. Try it yourself. It will not drain until air enters the top but you have to pull the pipe straight up out of the bucket of water.

Now on to the solar panels. For a moment, let's assume that the pump has just shut off but the instant before the vacuum release valve has opened so there is no dynamic head and only static head. Pressure decreases with elevation so at the lower end of the roof near the gutters, let's say it is 11.5' above the water level, is at -5 PSI. The pressure at the top of the roof, let's say again it is at 23' above water level, is -10 PSI. So the starting pressure at the very top of the roof is less than the lower part of the roof. Hopefully you can see at least this part. By the way, I have measured the pressure at the top of my solar panels and indeed as predicted by theory, the pressure is lower by the height of the building divided by 2.31. I always like to see theory in action.

Now, at this stage introduce the vacuum release valve. Now, the PSI at the lower part of the roof is at 0 PSI so the weight of the water will pull the water out of the ower pipe. The bottom of the roof is at 0 PSI but because the top of the roof is higher and it is still a closed system no air at the top of the panels yet, the pressure is still at -10 PSI due to the water weight on the return side. So there is still a negative pressure at the top of the roof as so water will flow up the panels to the top of the roof and out the return.

You can prove this concept to yourself with a small amount of hose and bucket of water. Clear hose works the best since you can see that it is full of water and you can see which way it flows out. Hold the hose completely under water until it fills. Bend in a loop such that you keep both ends in the water but the loop extends above the water. No water will flow out at this stage since air cannot get it. Place your thumb over one end and slowly pull it out of the water but make sure you keep the other end in the water. The end in the water is the return side for the solar and the end with your thumb is the vacuum release which has not opened yet. So you should have a tube filled with water and one end in the water and the other end with your thumb over the end and the center of the tube higher than your thumb. Now release the end of the tube with your thumb and no water will come out that end because it drained up and over the loop back into the water through the return side.

Hopefully this is much clear now and you have a way of proving it to yourself.

 

[duraleigh]

Ka-Zing!! I have no further questions! :lol: :lol:

 

[Sabot]

Great job Mark, made my mind go fuzzy again...

I just love that animation!

 

[JasonLion]

I wouldn't extrapolate too far from these simple models. The movement of air and water inside the pipe is rather more complex when you have a capped pipe with a hole half way down it's side, which is more similar to the situation in question. In that situation air can travel up the pipe and water down over a range of pressures. Likewise the simple capped pipe will behave differently when the pipe is tilted, as many of the pipes in solar systems normally are. Things get even more complex when networks of pipes, as occur in the panel, are involved. It is quite possible for air to be flowing up some tubes and water down other tubes under some conditions.

Regardless of these complexities, mas985 makes a solid argument that there will be negative pressure on the solar tubing when draining, regardless of the location of the vacuum breaker, which will apply force towards collapsing the tubes. It also makes sense that the negative pressure will be greater when the vacuum breaker is somewhere other than the top of the whole system. However, I have no reason to believe that this negative pressure will cause significant wear on the tubing. Many systems are installed with the vacuum breaker lower down and these systems observably do not fail dramatically sooner than ones with the valve at the top.

I think aspects of this discussion actually support my reasoning that the vacuum breaker is not put at the top so that it will get enough pressure to insure that it closes and stays closed. The entire solar system will work if there is even a slight amount of positive pressure at the very top of the system. But vacuum breaker valves can be prone to leaks at very low positive pressures. Many systems will have enough pressure at the top to close the valve reliably, but some will not. Placing the vacuum breaker lower down insures that if the system works at all there will be sufficient pressure to reliably close the vacuum breaker valve.

 

[Belldiver]

Is there any reason not to install one at the top AND the bottom? Seems this would solve all problems, so long as your pump puts out enough pressure to keep them fully closed.

 

[mas985]

Jason,

I didn't mean to imply that having the breaker on the bottom would not work or would definitely cause a problem. Life is full of probabilities and being the cautious person that I am, I would simply choose the top over the bottom. I think the risk is minimal and would probably only be an issue for poorly made panels. I was simply interested in the reasoning for putting on the bottom.

However, your last point is a very good one and worth considering if you have a variable speed pump with solar. Having the vacuum breaker on the bottom would allow you to operate at a slower speed than at the top. But usually, you want to run at fairly high flow rates (> 45 GPM) when using solar so having the break at the bottom may or maynot help depending on the situation but it would at least give you the option. Thanks for pointing that out.

[EDIT] - After thinking about this some more, Powermat warns about running panels under negative pressure so if you are placing the vacuum break at the bottom of the panel because the top does not have enough pressure, then part of the panels may be running under a slight vacuum. This is probably not a very efficient way to operate the panels and probably why Powermat and others recommend against it. If you are using EPDM, then there would be a tendancy for the tubes near the top to collapse and restrict the flow somewhat. Anyway, it is probably not a condition that I would want to operate my panels under.


A few more points on my post above:

The two pipe thought experiment was meant to explain static head and not be a proxy for the solar panels. However, the bucket and tube test was meant to be a proxy for the solar panel situation. The tube is about the size of a solar panel tube, the end of the tube in the water represents the return side of the solar panel plumbing, the bucket is the pool, the thumb over the other end of the tube out of the water represents the vacuum release and the upward curve of the tube represents the height of the panel. The length and height of the tube only changes the forces involved not the principles of siphoning. Since the panels are higher off the ground, the siphoning effects would be much strong and so the flow rates much faster when the panel drains out. But the principles of siphoning still apply.

One additional test that expands on the tub and bucket is to place one end of the tube in a bucket of water, fill the tube full of water and place a thumb over the other end of tube, move the tube with thumb over the edge of the bucket and below the water line. When you release the end of the tube a siphon is created and water flows up the tube over the edge of the bucket and down the other side of the tube. While water is flowing, raise the end where the water is coming out to the water level of the bucket but still outside the bucket. The flow rate will slow and then when the tube end is raised above the water level, the flow will reverse and the tube will drain back into the bucket. Again, the last part is similar to what is happening in the panels. Also, this is a great way to teach kids about physics.

 

[mas985]

Is there any reason not to install one at the top AND the bottom? Seems this would solve all problems, so long as your pump puts out enough pressure to keep them fully closed.

What would happen in that case is that the lower one would stay closed due to the weight of the water until the upper panels drained then it would open so it would be no different than having only one at the top.

 

[mas985]

I have a few more thoughts that I would like to share. I really love these types of discussion/debates as they seem to the collective understanding of complex issues. In addition, I know they help me to clarify my understanding about things that I have not previously put much thought into such as vacuum breaker location.

Anyway, thinking about this some more and some of you probably already know this but in theory at least, the vacuum release valve could be placed anywhere above the pool water line and the panels would still drain. This includes 1" above the water line although not very practical. However, water will always drain in opposite directions from the vacuum break no matter where it is. So if it is 1" above the water line, then only the water below the valve will travel downwards and all of the water above the valve will travel first upwards through the pipe, then through the panels and then back down the other pipe. Not very efficient but it would still work, albeit much slower drain time. As long as there is a small pressure differential between the two sides of pipe which is caused by the height of the vacuum valve, then water will always flow away from the valve in opposite directions. Doesn't mater the size of pipe or the height (up to 32') it goes to or even if solar panels are involved, this is always the case for two vertical pipes connected at the top (solar or not) and filled with water. However, syphons will break if water is lifted more than 32' because the water will start to vaporize and break the syphon.

Of course placing the valve below the water line would simply drain the pool.

 

[dschlic1]

On my solar system, the relief valve is placed about 3 feet above the pump. It works in that when the pumps stops, all of the water drains from the solar panels. My solar panels are made from HDPE which is a rather "hard" plastic, so under moderate amounts of occasional vacuum will not collapse.

However another issue is that the entire system, including the solar panels is under positive pressure when the pump is running. This positive pressure is greater than the pressure needed to pump water to the highest point in the system. Otherwise the panels would not fill. So while the pump is running, the panels are under a positive pressure. How much is determined by the hydraulics of the pool return.

 

[mas985]

Mark,

Lets look at this a different way. Your example at the very end is not representative of what happens in a solar system. You are creating a system that is open only at either end. In a solar system with a breaker, you have a system that is open at 3 points - the return, the source AND the breaker. Take your example and change it. Create a siphon, but put a hole in the siphon hose somewhere below the top of your loop. Cover the hole and create your siphon. Now, open the hole (both ends remain open as well). The siphon will stop and water will drain back into both ends. If the hole is below the top point in the siphon, some BUT NOT ALL of the water above the hole will be pulled up and over the top. Depending on the size of the hole, more or less water will be pulled up through the top - I contend a lot less than you think.

My example is exactly what happens with the solar system. It is effectively a siphon. With the solar, the only thing that is exposed to air is the pool and the vacuum break. If both ends of a line are in water, there is no way for air to get into the lines. When the vacuum break opens, there is only one way for air to get into either line and that is at the vacuum break. The upper portion of the panels will siphon water out of the panels and down into the pool (remember there is no air in this side). The lower portion will simply drain to the pool. In both cases, air replaces the water and water will drain away from the break point. Remember solar and pool plumbing are effectively a closed system. The pool surface is the only thing open to the air but it doesn't change the hydraulics any it only sets the reference PSI at the pool surface to 0 PSI.

If what you are saying is true, then if I close the end opposite of where the breaker is, then open the breaker, only the water below the breaker will drain, but the remainder (above the breaker) will remain in the pipes no matter how large the hole is. This isn't what will happen. All of the water on the open/breaker end will drain out.

This is exactly what will happen. There is no way for the water to escape. This is much like the straw scenario. Water will not drain out of the end of the straw if one end is plugged. Or if you like, use the tube example. Fill the tube will water, plug both ends make a loop and release the end that is higher (Edit: actually it doesn't matter which end is higher or lower. If one end it plugged, no water will flow out). Water will not drain out of the other end that is open to the air and higher than the other end. Try it. One of those bendy straws would work as well.

Another example. Using your soda straw example but going further, cut a hole above the waterline and cover it. Drink to fill the straw and cover the end and lift it out of the water. Now, open the hole. All of the water drains out of the straw, below AND above the hole. Your example is like lifting the straw out of the water while keeping the end covered. None of the water escapes! Negative pressure at work. My example is like a normal solar system - all of the water drains out, both below AND above the hole.

As for my tube test, I think you missunderstood the analogy. The open end was simply one half of the setup, only the upper panels and return drain tube. You can separate them because with vacuum break open, the PSI at the vacuum break is 0 PSI and thus each half can be treated separately.

So I have one more setup I think that you will agree is very similar to the solar panel setup and along the same line as what you had suggested. Take two pieces of clear vinyl tubing and put an irrigation tee (small barbed between them). Fill both tubes and place thumb over the tee outlet to prevent air into the tubes. Keep both ends of the tube in the water, make an upward loop with the tube and put the tee somewhere below the top of the loop and above the water line. So the tee is the vacuum break, both tube ends represent the suction and return lines. Let go of the tee and you will see the water separate and travel down/up each opposite tube away from the tee. It doesn't matter how high the tee is either. It could be 1" above the water surface and the same thing will happen. However, the closer you get to the water surface, the slower the flow rate is though the longer tube. Hopefully this test is more to your liking and will convince you as what actually happens in the panels. Please try this.

I would like to point out that this is not "my" theory and is predicted by physics and hydraulics. If you still don't believe me, you can confirm everything I have told you to online sources by googling "siphon". Also, I challenge you to run the tests that I have laid out and see for yourself. They are difficult to do on a solar panel itself but like most other physics problems, then can be scaled down with the same results.

Wikipedia description:

Theory
Liquids can rise over the crest of a siphon because they are pushed by atmospheric pressure. Siphons must be started by filling them in any number of ways. After priming, atmospheric pressure acts on both ends of the siphon, but the longer leg carries a greater weight of liquid. Gravity then drains the liquid through the longer leg, and this maintains the low pressure that was established at the start. Capillary action can enhance the siphon and cavitation may modify the phenomenon and cause the siphon to 'break'.[5].

Once started, a siphon requires no additional energy to keep the liquid flowing up and out of the reservoir. The siphone will pull the liquid out of the reservoir until the level falls below the intake or outlet of the siphon, whichever comes first. Energy is conserved because the ultimate drain point is lower than the liquid level of the reservoir.

The maximum height of the crest is limited by atmospheric pressure, the density of the liquid, and its vapour pressure. When the pressure exerted by the weight of the liquid equals that of atmospheric pressure, a vacuum will form at the high point and the siphon effect will end. The liquid may boil briefly until the vacuum is filled with the liquid's vapour pressure. For water at standard atmospheric pressure, the maximum siphon height is approximately 10 m (33 feet); for mercury it is 76 cm (30 inches).[dubious – discuss]

An analogy
An analogy to understand siphons is to imagine a long, frictionless train extending from a plain, up a hill and then down the hill into a valley below the plain. So long as the valley is below the plain, the part of the train on the valley side of the hill will be longer and heavier than the part on the plain side of the hill, so the portion of the train sliding into the valley can pull the rest of the train up the hill and into the valley. What is not obvious is what holds the train together when the train is a liquid in a tube. In this analogy, atmospheric pressure holds the train together. Once the force of gravity on the couplings between the cars of the train going up the hill exceeds that of atmospheric pressure, the coupling breaks and the train falls apart. The train analogy is demonstrated in a "siphon-chain model" [6] where a long chain on a pulley flows between two beakers.

 

[salinda]

Mark,

Bumping this--I need another clarification.

There is so much wrong with my solar system that I have had now for almost 5 years. The original installer screwed up: C-loops, coils on sloped roof, bank returns don't merge ideally, mix of coils and mats need different flow rates that are too hard to adjust. The I had a "pool optimization" done a couple of years ago by a company that is no longer in business. They listened to my explanations of what didn't work with the solar and assured me that the new variable speed pump (Ikeric) that they installed would work with solar just fine. Then they hooked the solar up backwards and I should have sent them packing and sued them. I didn't. I gave them a chance to fix it and they offered to fix the solar problems I was having but I think they just screwed it up worse. The solar does not work as it did before--I think the coils just need a high-flow wasteful pump. I got taken and it still burns me every time I think about it.

One of the things they did was to install an extra vacuum breaker so I now have 2 or 3 (I don't even know at this point). What I can tell you is that the mats do not drain completely when the solar valve turns off. I know this because when I drain the mats at the end of the season, hot water comes out of the hose bib that serves as my shutdown drain. At this point, solar has not been running for days or weeks already. This is stagnant water on my roof at all times.

At this point, we are waiting for gross system failure and then we will redo the whole solar system on the roof. As an intermediate step, I might abandon the coils at some point and just work with the mats, but they are not big enough for my large pool.

My question on this is:
If you have two breakers that are each near the top of the roof, won't water be trapped in any loops that are between them once the water level drops below the level of both vacuum breakers? Is that one of the things that is happening with my pool?

If so, maybe I can remove all but one of the breakers and solve part of the problems.

 

[mas985]

Salinda,

You are correct in that having two vacuum breakers could possibly cause water to be trapped in a section of pipe depending on where they are and how the panels are plumbed. In most cases, you want to stick with a single breaker. However, if your panels are plumbed in parallel and there are separate runs to the roof for each panel, then you want a breaker for each panel since they will drain independently. But that is probably the only situation where two breakers is a good idea.

However, even with a single breaker, it doesn't guarentee that all of the water will drain out of the panels. If there is any pipe route which creates a trap, then you could still have some water left in it. A single breaker helps to create a strong suction on both sides of the break and will remove most of the water out of the trap but a little will always be left behind. With two breakers and a trap in the middle, very little of the water will drain from the trap.

 

[salinda]

Hopefully there is no harm done just leaving it the way it is....is there any problem with water trapped in there during the non-freezing months?

 

[mas985]

Freezing is the only major issue with leaving water in the panels. However, it is suppose to get below freezing this week in the Bay Area.

 

[salinda]

I drained the panels manual a few weeks ago. Let's hope that they are tilted properly. I had no leaks this summer, even after last year's deep freeze, so I think it should be okay.

 

[el_steak]

If you have a 3-way valve to divert the water to your panels, make sure it is "leaky". If it is of the "Positive Seal" type, a lot of water will remain in your panel even if you have a vacuum break valve.

You can "fix" your 3-way valve by opening it up and drilling a small hole (3/8 - 1/2) inside.

 

[launboy]

I realize this topic is kinda old, but someone else brought it to the top of the stack, so I figure I'll post too.

Although everything said about siphoning water is true, and the math may work out, it just doesn't happen that way in real life(at least in my experiences). When you open a vacuum breaker that's not at the highest point, some water may flow upward, but (especially in larger dia. pipes) most of the time, the air just bubbles up past the water and the water just goes down instead of being siphoned up.

Just my 2 cents,
Adam

 

[salinda]

When you open a vacuum breaker that's not at the highest point, some water may flow upward, but (especially in larger dia. pipes) most of the time, the air just bubbles up past the water and the water just goes down instead of being siphoned up.

This must be what's going on in my case. Thanks for the explanation.

In answer to the previous poster, I do have a 3-way diverter that allows drainback. I also have a small bypass that would allow some drainage.

 

[mas985]

Although everything said about siphoning water is true, and the math may work out, it just doesn't happen that way in real life(at least in my experiences). When you open a vacuum breaker that's not at the highest point, some water may flow upward, but (especially in larger dia. pipes) most of the time, the air just bubbles up past the water and the water just goes down instead of being siphoned up.

Just my 2 cents,
Adam

I never meant to imply that the siphon process would ever remove 100% of the water from the plumbing. There are a lot of dynamics going on so some of the water will always be left behind. It is just a question of how much.

First off, the panels themselves have very small diameter tubes so it is very unlikely that the air will be able pass the water in such a small space. Cohesion and capillary action will prevent this from happening so nearly all of the water should be drained out of the panels. This one is easy to prove to yourself with a short piece of clear tubing while playing in the pool.

However, as you correctly pointed out this may not be completely true of larger diameter pipes that are forced to drain upwards. Here it depends on the rate of the water draining from the pipes as to how much water will be left behind. The air will take some time to displace the water as it travels up the pipe and as it displaces the water, that portion of the water will fall back down the pipe. However, the travel rate of the air is much slower than that of the water so the amount of water that falls back down should be quite small but again, it depends on the flow rate of the water and the diameter of the pipe. The higher the elevation of the panels, the faster the water will drain from the panels and the less water that will be left behind. In most cases, most of the water will be siphoned out leaving only a small amount behind.

But again, this problem can be avoided if the vacuum release is placed at the top most point of the plumbing. This way all of the water will flow downward and so it will be less likely for water to be left behind. Although, I have seen panel installations with intermediate high points which can always present problems and be very difficult to drain completely.