Solar series vs parallel

The heliocol manual shows series is acceptable up to the max number of panels which is 10 for the HC-40. If setup with the right flow rate, the water will have a relatively low temp differential so you should be fine. The tilt setup is important for both drainage as well as to ensure that the tubes all fill on each cycle (assuming that you have an automation controller)
I believe @Dirk has some good charts showing the importance of this.

rporonga said:

Thank you for replying to this thread. My understanding of what it says in the manual is that the max of 10 is for the parallel connection. Series can be used for large installations with more than 10 panels to break up the parallel groups of up to 10 each. My understanding is that series of the 2 groups will have some efficiency loss, but don't know how much would be lost.
If the total flow rate is the same, the series connection would double the flow in each panel, so the temp increase in the first row would be about half, and the total temp increase would be about the same. The head loss would double in the series connection of 2 rows, but I think this is just 1/4 foot per row. Is that low enough not to waste too much extra power at the pump?
How can I check that the low side is the feed after the installation is complete? I don't see any eay for checking in which direction the water is flowing.

Click to expand...

Just follow the pipe from the system up to the panels. It will always enter at the low end of one of them.

rporonga said:

Thank you for replying to this thread. My understanding of what it says in the manual is that the max of 19 is for the parallel connection. Series can be used for large installations with more than 10 panels to break up the parallel groups of up to 10 each. My understanding is that series of the 2 groups will have some efficiency loss, but don't know how much would be lost. If the total flow rate is the same, the series connection would double the flow in each panel, so the temp increase in the first row would be about half, and the total temp increase would be about the same.

Click to expand...

That is correct and if you look at most efficiency charts for heat loss vs GPM, halving the flow rate decreases efficiency by about 12.5% (80% to 70%). So that is not a big deal.


However, when you have too many panels in parallel, the water flow may not get to the end panels even if the supply and returns are on the diagonal. This is why most manufactures have parallel panel limits. Beyond this you are losing efficiency also due to lower flow rates on the end panels.

But there is another solution. If you split your supply pipe and return pipe into two paths/sections, you can set up the panels in 2 parallel sets of 5 parallel panels each. Since each set has only 5 parallel panels, pressure loss is not an issue. Also, as long as the two sets are at the same elevation and the supply and return pipes are the same size and length after/before the split, flow rates should be identical between the two sets of panels.


This is more energy efficient for both the pump and heat transfer.

rporonga said:

The head loss would double in the series connection of 2 rows, but I think this is just 1/4 foot per row. Is that low enough not to waste too much extra power at the pump?

Click to expand...

If you have 2 panels in series, the difference in head loss is insignificant. Remember, you are targeting a specific flow rate per panel (~ 0.1 GPM/sq-ft). This is very low from a flow and head loss perspective (~0.14ft) so most panels should not experience much in the way of head loss unless they are overdriven.

rporonga said:

How can I check that the low side is the feed after the installation is complete? I don't see any eay for checking in which direction the water is flowing.

Click to expand...

You should be able to remove an end cap.

In all cases, defer to @mas985! That being said, this is what I think I know.

There is a max number of panels you're supposed to connect together into a single array. It was 8 for my 12.5 footers, perhaps the number is higher for smaller panels. Regardless, if you want/need more panels than the max for one array, then you make two arrays. Simple enough. But what you don't do is run the output of the first array into the input of the second array. If that is what they did, then I would agree that it needs to be replumbed.

What I understand to be the correct way, is that you split the supply side into two pipes, then each of those pipes feeds the input of each array. Then the outputs of each array join back together, then on to the pool. I believe that is exactly what Mark's drawing illustrates.

So basically, each array heats its "share" of the water equally.

The second thing I think I know is that that panels should be inclined in two different directions. The obvious one is that the supply pipe is at the bottom of an incline, and the return pipe is at the top. That's so the panels can drain properly when they are done for the day. But the arguably more important tilt is from the input corner to the output corner (which should be on opposite sides of the array). So like lower left to upper right. That tilt, along with the incline, forces the water to completely fill the array, without trapping air in one of the upper corners, which would really degrade the heating efficiency. This page explains these concepts better than I can:


I built my own system and followed that website's advice to the letter. Seems to work! But as always, I defer to Mark to punch holes in anything I have misunderstood!

1.
I think it's possible that sun on the roof-top temperature sensor could signal the solar controller to start heating before the sun actually heated up the water leftover in the array. Which means when the pump fires up, instead of pushing a "new" batch of pool water onto the roof and displacing the air that filled the array the day before, it'd be pumping "leftover roof water" right into your pool that has been cooling off all night, certainly a lot colder than the pool water. I can't say how that would affect the overall heating efficiency that day, but it certainly is bad form.

2.
You shouldn't have to remember to drain your array every time a freeze is imminent, or even remotely possible. What if you aren't even home? Or forget? Or otherwise just can't? Unless your arrays are below the pool's water level, plumbing them so that they drain themselves automatically each day (regardless of which State they live in), is so easy to do as to make it pretty much ridiculous not to. (IMO)

3.
There may be other reasons why it's good practice that the array(s) are emptied each night that I'm not thinking of (@mas985?).

Are you saying the source water of your array enters the array into the top manifold (the highest point of the array), and then exits the array from the bottom manifold (the lowest point)? I can't say exactly why that is bad, but as far as I know that is never done, so there must be a good reason for it. @mas985? That can't be right. I think that setup could easily trap air somewhere in the array, and any air at all in the tubing is a loss of heating efficiency.

And while it's hard to say for sure without pics, if there is a way to incorrectly put water into the top of your array and take it out at the bottom, then certainly there must be a way to do it correctly. Maybe those pipes can't now be crossed up on the roof, but certainly at some point lower!

Pardon for saying, but solar installers that don't know enough to feed an array from the low point, while also running arrays end-to-end, sure sound to me like they have no idea what they're doing, or worse, just don't care enough to do it correctly. So odd.

Regarding the tilt I described, even the solar company that sold me my array didn't understand that concept, and said they never do that. So it doesn't surprise me that it's not in the Heliocol installation guide. But if you watched the animations I linked, it sure seems pretty clear to me why this is pretty important.

A solar heating system is not going to heat your pool to excess. It barely can make it a little warmer. There is no wiggle room, nor abundant amount of extra heat to be wasted. You need every therm. This is the one point where @mas985 and I are not 100% in sync. He sometimes states plumbing and flow rate scenarios, that might short you a small percentage of heating efficiency, as no big deal, something that would not be noticeable to the user. And I can't argue that's not true. But I'm running my pump and I paid for the panels, so if I can just as easily eke out a little more efficiency as simply as not, then why wouldn't I?

Full disclosure: I have PV solar so pay nothing for pump runtime or extra flow rate. @mas985's MO (if I may) is addressing the concern of heating efficiency vs the cost of the electrical energy to provide that heating. Since I have no cost of electrical energy, that makes me lean on the side of heating efficiency. As long as you understand the principals, then you can make up your own mind about that.

Mark, pardon me if I'm incorrectly interpreting your thoughts about such things.

Dirk said:

What I understand to be the correct way, is that you split the supply side into two pipes, then each of those pipes feeds the input of each array. Then the outputs of each array join back together, then on to the pool. I believe that is exactly what Mark's drawing illustrates.

Click to expand...

I didn't describe one important point. Where the two return pipes from the two arrays meet is important. They should meet at the highest point in the system. If you run the return pipe of one of the arrays into the other array's return pipe somewhere lower than the highest point, it's possible one of the arrays will never fill correctly. The site I linked describes that.

I see the series installations in the manual. I cannot explain why this is offered. It contradicts what I know to be true (which doesn't mean much, I admit). From the manual:

COLLECTOR LAYOUT & PLUMBING
Collectors can be installed in a multitude of configurations and methods, but systems should adhere to a few key rules for optimal performance.
1. Systems that are pitched should be installed with the feed line(s) entering the collector bank(s) on the lower manifold and the return lines exiting the banks(s) on the opposite, upper manifold end. Systems that are on low slope roofs or otherwise “flat” surfaces may should
still ensure that the feed and return on are opposite manifold ends from each other.

Click to expand...

That solves for any argument about that from your installer. Unless he thinks your slope is too little too matter. But if there is any slope, then I would say it does matter.

2. Systems should be plumbed in a reverse-return fashion whenever possible. This means that the return to the pool occurs as high up and/or as far away from the feed as possible to ensure even, full flow through the system. If reverse-return plumbing is not feasible it is important to use balancing valves through the system to ensure balanced flow to all collectors and avoid short-circuiting any point in the system.

Click to expand...

I could argue this refers to the tilt I described. Or at least that the tilt satisfies this "rule." I believe the key phrase is "even, full flow." If the array is tilted the wrong way, then it's going to trap air in one of its corners. That is just physics. And as I said, if there is any air in the system, you're losing heating efficiency. I suppose if the array is perfectly parallel to earth, then it will clear all the air. Which I guess is what installer rely on. By putting a small tilt in my array, I don't have to worry about how perfect my roof is. It simply cannot trap air. If nothing else, it's just an extra added measure to ensure "even, full flow."

3. Systems should always have a means to drain down from the roof, preferably automatically by gravity when the pump shuts off. For systems that cannot automatically drain a means of manual drainage should be installed as required. Never create a situation where water can be trapped in the system. Large diameter piping (2” and above) will drain even if perfectly horizontal so there is no need to further slope horizontal piping runs.

Click to expand...

That's pretty clear. The exception is when the array is below the level of the surface of the pool. That's when you need a manual drain. I would not read this as "put in a faucet if you're too lazy to create proper automatic drainage!"

Side bar: when I bought this house, it had two ridiculous back yard patio overhangs. The kind that are just 2x2s six inches apart, providing no real shade or rain protection. The first thing I did was slap some metal corrugated roof on 'em. But they were built flat, as originally there was no need to provide a slope for drainage. So I solved that by lopping off a few inches from the bottoms of the two posts supporting the structure. I jacked up one corner, cut off the bottom of the post, and then dropped it down. Then the other corner. Yah, it was bolted and nailed to the house, but the slope I created did virtually nothing to that connection. You only need 1/4" per 12", so with a 12' patio, I only had to lop off about 3" or so. Works great.

Would it be possible to add some slope to your patio cover in this way, to improve (or create) the drainage your panels should ideally have?

rporonga said:

I do pay for power, so I'd rather use the configuration that is most power efficient. It is not clear to me if power and heat efficiency are fully correlated or not. If the heat efficiency is higher then I can run the pump slower or for less hours, right? Anyway the parallel vs series is a battle I may not win.

Click to expand...

Ha, ha. Mark and I have debated this before. The more water you pump through a solar array, the better the heating efficiency. Up until a point of diminishing returns. You either bump up against the amount of pressure the array can handle, or the cost of pumping more water just isn't worth the small improvement of heating you get.

Heliocol recommends an optimum flow rate for heating efficiency. I've assumed they take the above into consideration. So because I don't care about electrical costs, I have my pump set to meet that optimum flow rate.

But Mark contends that you can run the pump at a lower flow rate, and save on electrical costs, without noticeably impacting what it feels like in your pool. He agrees you'll lose some heating efficiency, but that you won't notice the difference. I am hard pressed to argue that. It is sound thinking. But as I mentioned, I don't need to save money on electricity, I want my pool to get as warm as possible (whether I think I can feel it or not).

So you will have to determine for yourself which way you want to go. Or try out various flow rates (pump RPMs) and see if it makes a difference in how the pool feels and how the electrical bill looks.

So yes, power consumption and heating efficiency are definitely correlated.

The most efficient way to heat a pool with solar is to also have PV solar. And PG&E is just going to continue to raise rates, so that statement will only get more true over time. Unless you see yourself moving in less than five years or so, the PV will pay for itself. Mine has, and now I truly get free electricity. So I can run my pump when I want to, and skim on high, or pump water up onto my roof, or run my vac all I want to, without having to think about energy costs.

As a bonus, when I got it, I got a huge break on my income tax for several years to boot! I'm not sure how the current cost of solar plays out, but it, too, is only going to get more expensive and be less of a good buy. The best day to get into PV is today, tomorrow will be a little worse, the day after worse still. That is PV-solar-salesman-speak, which I fell for, so I choose to believe it!

rporonga said:

I do pay for power, so I'd rather use the configuration that is most power efficient. It is not clear to me if power and heat efficiency are fully correlated or not. If the heat efficiency is higher then I can run the pump slower or for less hours, right? Anyway the parallel vs series is a battle I may not win.

Click to expand...

Power and heat efficiency are negatively correlated. Faster flow rates mean higher heat efficiency but lower power efficiency and visa versa. However, as I pointed out earlier, a 50% reduction in flow rate per panel, decreases heat efficiency by 12.5%. However, pump power is reduced by almost 85%. It just depends on what is more important to you. In CA with tier 4 over $0.40/kwh, you know what my choice would be.

As long as the panels can generate more heat than is actually needed, the loss in heat efficiency can easily be made up for in longer run times and it is still more energy efficient. I suspect with 10 panels, you won't be short on heat.


Also, is this the way the panels are currently plumbed?



I am still not 100% clear from your description. Any chance you can get some pics of the roof?

One more thing, as long as you target 0.1 GPM/sq-ft of total panel area, the total efficiency difference of the series vs parallel becomes much less significant. In the above picture, the per individual 4x10 panel flow rate would actually be 2x higher than in the configuration below in order to meet the 0.1 GPM/sq-ft. If you were to look at the heat distribution of both methods, you would find that the average panel temperature across all panels is actually the same. It is only distributed differently. So technically, there should not be a huge difference in efficiency.

It's pretty difficult, for me anyway, to picture all this from your text. A drawing or photo would make this conversation so much easier.

It is my understanding that to calculate total optimal flow rate, you multiply the optimum rate for one panel times the number of panels. My panels require 5 GPM (each), times 8 panels equals 40 GPM. I don't see how how they are connected would change that math. I think Mark offers another formula, based on square footage of the panels, but again, I don't see how his formula would change based on series or parallel.

Less flow = less heating efficiency but less pump energy required. Also don't see how parallel vs series would change that either.

I didn't want to approximate total flow rate, or even take the time to learn how to approximate it, so I installed a FlowVis flow meter. Dialing in the 40 GPM I need is very simple, and I can monitor and adjust that throughout the year as it varies (as my filter gets dirty, my flow rate drops, so I can check my meter and goose the RPMs to compensate as needed). Mark might know other ways to figure out the flow rate through your plumbing.

A VRV doesn't have to be at the highest point. Mine is below the lowest point, in fact, just under the eve of my roof. But it's location can effect the RPMs (flow rate) required to close it. I gotta defer to Mark on that, too. I've read his explanation of that many times and I still can't quite wrap my head around it. I just know my VRV works where it is. I chose to put it under the eve for two reasons: to keep it out of the sun to increase its lifespan, and to make it easy to replace, if need be, without having to climb around on my roof tiles.

If I use Mark's formula on my panels, I get:
12.5' x 4' = 50 sq ft x 0.1 = 5 GPM per panel
Exactly what Heliocol said it should be.

Times 8 panels = 40 GPM.
For reference: I need 2200 RPM to get 40 GPM of flow through my array (with a clean filter).

rporonga said:

Thank you for all the explanations. The series vs parallel argument on efficiency makes sense to me. The only diffetence being the slight increase in pressure loss from the series connection plus the extra loss of doubling the flow per panel, correct?

Click to expand...

Yes but depending on the flow rate through each setup.

rporonga said:

The installer claims that with the series connection the pump can be run slower and hence save power. Maybe he has a point then?
With parallel I would need 40 gal/min. With series I could go as low as 20 gal/min and still get 4 gal/min per panel.

Click to expand...

Only if you are targeting a 4 GPM per 4x10 panel. However, this also means that the series version is running at 15% less efficiency that way because the inlet to outlet temperature is twice as high than the parallel case.

My suggestion is to run 40 GPM total for either series or parallel so their inlet/outlet temperature difference is the same and their efficiencies are the same. In this case, the series version will have higher head loss so the RPM will need to be higher in order to reach 40 GPM and thus require more pump energy to run.

rporonga said:

He set the pump to 2950 rpm, 2.7 hp pump, 10 feet high roof, 100' pipes. Any clue of what flow I may have?

Click to expand...

I can give you an estimate but I would need to know the rest of the plumbing layout. But what pump model do you have? If it is an Intelliflo, the menu should show the flow rate.

rporonga said:

I'll try to snap a pic tomoorrow. This is on a 21'×24' patio cover slightly tilted. The panels are arranged with the headers horizontal, with both groups of 5 panels at slightly different height due to the roof tilt.
The vacumm release valve is in the middle, where the 2 groups are connected in series, not at the highest point. I don't know if this can create an issue with draining the top panels, as the vacuum breaker is below them.

Click to expand...

VRV can go pretty much anywhere. However, the closer to the pump it is, the lower the RPM you can set the pump to keep the VRV closed. So if don't you need high heating efficiency, you can sacrifice that to have lower pump energy.

rporonga said:

The installer is coming tomorrow to check the plumbing. Should I ask him to move the valve to the highest point? Could that make the valve leak due to the lower pressure?

Click to expand...

I would move it closer to the pump.

rporonga said:

To answer the question about the layout, yes they are connected as in fig 6.4 in your post, but the groups are top bottom and not side by side. Also, probably does not matter, but the feed and return lines are both at the right, so the left panel group in the figure is at the top and mirrored left/right in my case. So the series connection is a very short U. Without the mirroring, the series connection would be 20' of pipe in the top/bottom layout.

Click to expand...

Panels should always have the supply and returns on the diagonal. Doing a loop like that will reduce panel efficiency. The first panel will have the highest flow rate and each panel after that, successively less flow rate.

If you have an IR thermometer, you can actually get a feel of the efficiency by measuring each 4x10 panel temperature. Efficiency is the highest when all panels have the same temperature and the panel itself is uniform temperature. The greater the disparity in the panel temperature, the lower the efficiency.