Unexpectedly high system pressure when Solar is on

[rlab]

@mas985 I split this from No Chlorine Generation When Solar Heat On , as it was getting a bit off topic.

When my Solar is on, I am seeing an 8 PSI increase at the filter pressure gauge (from 4 PSI to 12 PSI). I have an Intelliflo VF which is set to run at a constant 25 gpm both with solar on and off. So the PSI increase is entirely down to added TDH from solar (rather than flow rate increase).

My back of the napkin calculations say I should only be seeing a 4 PSI increase, so it seems significantly over pressure.
- All 1.5 inch plumbing
- 50 feet of pipe length (not counting the solar headers)
- 25 by 90 elbows
- 6 by 45 elbows
This gives me a TDH for the plumbing of 2.8 PSI

The highest point on the panels is 12 feet above water line. The panels themselves are split into 2 arrays of 7 (14 panels total). The 2 arrays are connected in serial, so this gives 3.5gpm per panel. According to the manufacturer specs (Heliocol) this gives a TDH of ~ 0.18 feet per panel.
14 panels * 0.18 feet = 2.52 feet = 1.1 PSI

TOTAL PSI = 3.9, which is well below the 8 PSI I see in practice.

Are my calculations wrong, or is there something going on with the system?

 

[mas985]

What are the individual panel sizes?

If you double the flow rate, do you see any air coming out of the returns? This could indicate that the panels are not fully primed.

 

[rlab]

Each panel is 6.5ft x 4ft, so they are fairly small panels. And the fact that they are split into 2 separate arrays connected in serial means the net flow rate is low (7 * rated panel flow instead of 14 * rated panel flow).

- At rated speed (3.5 gpm/panel as per above), there is a minor amount of small bubbles from the returns. Barely noticeable.
- At high speed (5gpm + per panel) the bubbles go away completely
- If I run at high speed for a bit, then drop down to rated speed, there are no more bubbles.

Going high speed first then dropping to rated speed maybe reduces the pressure a small amount, like 0.5 psi (versus just running only at rated speed).

 

[mas985]

What was the filter pressure at 35 GPM?

Is the VRV after the second set of panels?

The problem may be the location of the VRV. If it is after the second set of panels, there may be too much pressure loss to keep it closed. If the VRV is not remaining closed, then the static head will remain.

 

[rlab]

Filter pressure was 18 psi at 35 gpm. It just goes up and up as I increase pump rpms.

VRV is on the supply side before both sets of panels. It is up on the roof at the lowest point of the panels.

The VRV seems to just be threaded in. Is it safe to unscrew it and screw in some kind of end cap instead just as a test to see if removing the VRV makes a difference?

 

[mas985]

If the VRV is on the supply side, then the pressure should keep it closed. Especially since the panels are only 12' high. The pressure at the VRV only needs to be slightly above 0 PSI so at the bottom of the supply pipe that would be about 5 PSI which is well below what the filter pressure shows.

But something in your solar setup is adding way too much head loss. That is 2-3x what is normally seen in a solar setup.

What is the history of this setup? Did you install the solar and has it always operated this way?

Cab you post pictures of your setup, pad and solar?

 

[rlab]

I will try to get some photos this weekend.

It was installed by a professional installer who is certified for this panel brand. Was only installed a couple of months ago and has always operated this way. I asked about the high pressure when it first went in, and was met with indifference by the installer. He said the majority of systems they install are on single speed pumps, so the pressure just is what it is, its not something they ever care about as long as the panels are producing heat.

One question in the meantime, if the VRV is operating correctly, then how are small bubbles getting into the system when running at the rated speed?

 

[mas985]

Are you sure those aren't swg bubbles?

If they are from the VRV, it isn't closed and could explain some of the excess pressure. But that would also mean a lot of head loss between the solar valve and the VRV.

 

[rlab]

Bubbles are definitely not from the SWG, I have tested in service mode with the SWG off and everything else off in the system, and still see the bubbles. They are only present when solar valve is on.

Where you say "That would mean alot of head loss between solar valve and VRV", why is that the case? i.e versus the head loss being somewhere else in the system?

If I temporarily unscrew the VRV and put a pressure gauge in its place, will that provide any useful information?

 

[mas985]

Where you say "That would mean alot of head loss between solar valve and VRV", why is that the case? i.e versus the head loss being somewhere else in the system?

In the 25 GPM case, the filter pressure is 14 PSI. In order for the VRV to open, the pressure has to be less than 0 PSI at the location of the VRV. So from the filter gauge to the VRV, there is a loss of 14 PSI or 32' of head. In order for that to happen at 25 GPM, there would need to be an equivalent of about 875' of 1.5" pipe.

If I temporarily unscrew the VRV and put a pressure gauge in its place, will that provide any useful information?

That would tell you for sure what the pressure at the VRV is and if it would open. But if the VRV is not at the highest elevation, since it is on the supply side, it would need to close in order for water to make it back out the returns. Otherwise water would leak out of the VRV and you would have no water out of the returns.

 

[rlab]

Thanks for all the help @mas985 , I have some testing to do this weekend and will see what I can figure out.

I'm assuming there is no magical device that lets me take non invasive pressure readings at various points in the system... I am limited to wherever there is a plumbing point I can screw a pressure gauge into.

 

[mas985]

Unfortunately, you have to penetrate the pipe to measure pressure unless there access another way.

 

[rlab]

Managed to test some things. First of all I removed the VRV completely, and this made no difference at all to the pressure or bubbles in the returns. So its definitely not a VRV issue.

Then I tried starting at panel rated flow speed, run for 30 minutes, then increase speed a little, wait 30 minutes, repeat. Did this all the way up to double the panel rated flow speed. I found exactly the same pattern each time:
- small amount of bubbles in returns
- each increase in pump speed results in a big rush of bubbles
- after 2 or 3 minutes this settles down back to small amount of bubbles

So I suspect my panels are just never priming. Each time I increase pump speed I flush out a bit more air, but never manage to fully prime until I hit double the rated speed (6gpm per panel), when bubbles seem to finally go away.

I suspect this may be related to my panels being installed in serial in 2 separate arrays, as per this diagram (in reality each array is on a different roof section). The installer indicated this is completely standard, and is the way they deal with panels split across multiple roof sections. But could that be the cause of failing to prime?

 

[mas985]

In series is not the best way to plumb panels. They should be plumbed in parallel. In parallel reduces head loss and increases thermal efficiency.

I think the problem you are running in to is that you need a minimum flow rate to force air downwards against it's tendency to rise in water. Plumbing in series just makes it worse. 50 GPM is not uncommon for a priming flow rate. When I had mine on a two story house (they are on the ground now), it took at least 40 GPM to prime the panels and purge all of the air.

Normally when someone wants to run panels on a VS pump at a lower speed, there are two criteria to be met. First the VRV has to be installed at a lower elevation than the panels. The height depends on the RPM and filter pressure. Second, the controller must be able to support priming at one speed and then stepping down to a lower speed. If the controller cannot do this, then you are stuck at a speed and flow rate required for priming.

 

[rlab]

OK unfortunately my Easytouch and Intelliflo do not have that option of setting a higher solar prime speed.

Other than added tdh, Is there an issue with just running the panels without full prime at their rated flow speed?

In my case running at rated flow without prime still uses much less energy than running at double that flow with prime.

 

[mas985]

Without fully priming the panels, there could be air in the panels and so they will not be adding heat.

 

[Rocket J Squirrel]

That seems like a strange layout to me. The VRV is really in line with the supply? I thought the purpose of the VRV was to allow air into the panels so they could drain when the solar disengages. That means one side has to be open to the air. When the water enters, water pressure closes the VRV. No high speed priming needed. This diagram adapted from yours shows relatively, not totally accurately, where my 2 VRVs are. Not inline with any plumbing, but at the ends of the panel arrays. And having the water exit from the highest point seems to defeat the purpose as well. Maybe things go the opposite way Down Under?

 

[rlab]

Water has to exit from the highest point, otherwise the panels don't fill unless you run them at a very high flow rate.

I wonder if adding a constriction to the return pipe is an option? Yes it would increase tdh, but i guess the hope is by letting the panels prime at a lower flow rate, it is less added tdh than having to run them at double speed.

 

[Rocket J Squirrel]

Water has to exit from the highest point, otherwise the panels don't fill unless you run them at a very high flow rate.

Ah, OK. I realize now a big difference between our setups is that my panels lay flat on top of a pergola, so there really is no high or low point. But still, it seems to me the VRV has to be open to air and not inline with the plumbing.

 

[Rocket J Squirrel]

it seems to me the VRV has to be open to air and not inline with the plumbing.

... or maybe I'm incorrectly assuming your VRV is like mine. Mine are single-ended with the outer side spring-loaded to the air. Maybe yours has a 3rd opening on the side to let air pass? I apologize if I'm taking this issue in the wrong direction.