The point of no pressure change.

Warmwater

Can anyone advise?


1) If there was no expansion vessel in a sealed system (just a pump and some pipe filled with water), where would the ponpc be then?

2) What if you installed 2 expansion vessels? Which one would be the ponpc?

3) Where is the ponpc on a circuit which is hydraulically separated from the primary circuit?

4) Where would the ponpc be if the expansion vessel was fitted to a parallel leg off the primary circuit. Like in a heating system, if someone fitted it to a flow pipe feeding a radiator?

Ironman

You must have an expansion tank in a closed system or else the relief valve open as soon as the temperature of the water rises.

With p/s piping, the secondary “sees” the entire primary (where the expansion tank connects) as the PONPC.

Jamie Hall

In your case 1, the question is quite meaningless. The pressure at any given point within the system would vary with flow and headloss in the system; if the pump were running, the lowest pressure would be within the intake volute of the pump, and the highest at some point in the pump diffuser, depending on the design of the pump. The pressure within the system relative to outside would be meaningless.

In cases 3 you are looking at an expansion tank on a circuit hydraulically separated from the primary circuit. If the two circuits are, in fact, hydraulically separated -- no flow connection between the two -- the expansion tank will influence only the circuit to which it is connected. The other circuit, which cannot exchange fluid with the first, won't be affected by the tank at all, and can have any pressure and pressure distribution it wanted. There will be a point of minimal pressure change in the circuit with the tank. The concept is irrelevant in the circuit without the tank.

It would be well to understand the "hydraulic separation" as the term is commonly used in heating systems is actually incorrect, unless the circuits are separated by a heat exchanger (they sometimes are). The common terminology, involving low loss headers or closely spaced Ts or what have you -- even large tanks -- does create flow separations which results in the flow in some circuits being quite independent, in terms of flow, from others -- but not in terms of pressure, since fluid can and does flow from one circuit to another, and at those points the pressure in the two circuits must be exactly the same.

At this point, too, it would be well to realise that an expansion tank is actually a point of minimal pressure change, not no pressure change. However, if the volume of the fluid in the system does not change then the pressure in the tank also does not change, and if there is only one tank the point of no pressure change will be where the tank is.

On to case 4. As stated above, he point of minimal pressure change will be at the tank connection, wherever than might be. Pressure downstream from the tank will be lower and pressure upstream will be higher. In the case of parallel circuits, the arithmetic can be a bit messy, but it is also important to remember two fundamentals: first, flow is always from higher pressure to lower pressure, and second, at any point where flow splits or rejoins, the pressure at that point in both flows must be the same.

Now to case 2. The point of minimal pressure change will be somewhere, considered in terms of flow, between the two tanks, and will depend on the flows in the various pipes connecting them, the gas and fluid volumes in the tanks, and their physical relationship to each other in space. The location will be highly unstable and, depending on a host of factors involving pipe sizing, fittings, and so on may even oscillate. In some extreme cases one or the other tank will completely non-functional and can be ignored.

Warmwater

Jamie, thanks for the in-depth response!

I'm not sure we're talking about quite the same things. I'm questioning the theory behind the point within a closed system when the dynamic pressure difference, created by the circulating pump, changes over from positive to negative. This is know as the "point of no pressure change" or "neutral point". At this point, the pressure is equal to the static head that the system started out at, before the pump runs. The pump has no influence at this point and the dynamic pressure is 0, as it is changing from positive ("pushing") to negative ("pulling"). This we know happens where the expansion vessel is connected and is why it's best to install the pump so it "pumps away" from this point, to maintain positive pressure through the system.

Case 1) Yes, I guess in reality the question is meaningless haha. Yes, I understand that the pressure difference created is with respect to the system losses and flow rate. I'm just wanting to understand at what point does it change from positive pressure to negative pressure within the system. I know it's where the expansion vessel is connected but, suppose the expansion vessel lost its air or was isolated from the system. Ignoring thermal expansion, if the pump was to run, where would it now change from positive to negative? See file attached for picture.

Case 2) I'm not sure on your response. Maybe I'm not explaining well. See file attached for picture.

Case 3) I take your comment about the term "hydraulic separation" really implies that there is no physical contact between the fluid on primary and secondary sides of a system haha. In the industry, we know however, that this is referring to something like a buffer tank or low loss header. So, where would the PONPC be on the secondary side if the expansion vessel is fitted on the primary side? My thought is at the buffer or low loss header. See attached file for picture.

Case 4) See attached picture. I know you would never do this, just interested in the theory.


Thanks again for the comments.

Luke

EdTheHeaterMan

This is very interesting and I understand your train of thought. (my train of thought missed the station long ago). This discussion begs the question, Who Cares?

Or, a more tactfully posed query, Why do you need to know?

I can see it in my mind's eye as the pressure might get to the point of equilibrium. It would be the center point of the closed-loop in relation to the circulator pump. Or better explained... after 50% of the friction loss of the loop is completed by the fluid in the loop. That is the center point in terms of friction loss of the components of the loop.

Here is a look at just one of your illustrations
With the pump off both tanks are at the static pressure of the system. With the pump on, the discharge pressure will compress the air in the upper tank. at the same time, an equal amount of pressure will be subtracted from the lower tank. These tanks cancel each other regarding the amount of air pressure is exerted on the system. the amount of air and the amount of water stays constant. the air compressed in the upper tank and the air in the lower tank is still the same by weight, and the combined volume is also still the same. the point of equilibrium is as if there was no tank at all, at the 50% mark in the loop.


Still, why does one need to know?

BTW @Jamie Hall great explanation!

EdTheHeaterMan

@

Warmwater said:

Jamie, thanks for the in-depth response!


Case 4) See attached picture. I know you would never do this, just interested in the theory.


Thanks again for the comments.

Luke

You would be surprised how many times Case 4 is out there in real basements across this world of ours!

hvacfreak2

Case 4 ( or something similar ) was typical pre - DH ( Dan ) when I started in the trades in the late 1980's.

Jamie Hall

Hmm... well, try to be a little clearer here.

First of all, one needs to be certain about what one is talking about with pressure. One can either consider pressure relative to some other pressure, or one can consider absolute pressure. One is almost always talking about relative pressure -- often termed "gauge" pressure -- and at that point, one needs to answer the question "relative to what?". If one plunks a pressure gauge onto a system, one will almost always be measuring gauge pressure, and equally almost always relative to the local atmospheric pressure.

When a physicist speaks of pressure, however, they are speaking of absolute pressure. Absolute pressure can never be less than zero (although in some highly sophisticated laboratory equipment or interstellar space it can come very close to zero).

So with that all in mind, let us consider your case 1, and define, quite arbitrarily, the reference pressure as the pressure existing in the system with the pump off. Now when we turn the pump on, the pump will create a pressure difference and a flow. With some consideration, it will be seen that since nothing has been added to the system (we're presuming the temperature stays constant) and nothing has been taken away, nor has the system volume changed, the average absolute pressure -- which in a dynamic system we must define as the total force exerted on the enclosure of the system divided by the area of that enclosure -- does not change. As you point out, at some point in the system, the local absolute pressure will be the same as it was before. Immediately after the pump it will be higher. At the pump intake it will be lower. This location is determined by the overall way the pressure in the system changes with flow; in a simple system consisting of a simple circle of uniform pipe, it will be half way around. In a more complex system it could be almost anywhere -- and indeed it is not difficult to envision a design in which there were multiple points where the absolute pressure was the same as the original pressure, with stretches in between where it was less and other stretches where it was more. In fact, one could design -- rather easily -- a system in which the minimum absolute pressure was less than the pressure at the intake of the pump. If one is thinking only of relative pressure and things pushing or pulling, this could become rather confusing.

Now let us consider the case where we have what the industry, if you please, terms hydraulic separation. It is absolutely critical to understand that in the industry terminology, what we have is not hydraulic separation, strictly speaking, but flow separation. This can be seen immediately if one realises that it is possible to drain such a system from a low point. Disregarding elevation effects, then, when the system pump or pumps is turned off, the absolute pressure throughout will be the same. We also have to introduce one other law: at any junction between two or more pipes or vessels, the absolute pressure in all of them at that point must be same, and we might as well introduce another law: the total flow into such a junction must exactly equal the total flow out. It makes no difference whether it is a half inch T or a half inch pipe entering a 10,000 gallon tank.

Now whether or not there is an expansion tank on the system, if we do not change the temperature of the contents or the size of the system in some way, we clearly have just a more complicated version of the first case, above. The average absolute pressure will not change, and the local dynamic pressure will be entirely a function of the various head losses or gains as we proceed around the system.

Now in a heating system, we have a complication: we are adding energy to the system in the form of heat. One of two things has to happen: either the volume of the contents must be allowed to increase, or the energy has to be absorbed by elastic energy stored in compressing the fluid and stretching the enclosure. In how water heating systems we are faced with a rather incompressible fluid and, generally speaking, with enclosures which are hard to expand -- so the absolute pressure will rise quite dramatically, even with small changes in temperature. It is much preferable, then, to allow the contents to expand more or less freely, and this is the purpose -- and only purpose -- of the expansion tank. In modern bladder tanks or older compression tanks, this is achieved by compressing the air in the tank; since air is much more elastic than water, the corresponding pressure change with volume change will be rather small. If our fluid cools, the air will expand to restore (nearly) the previous conditions. If it heats, the air will compress to again restore (nearly) the previous conditions. Now if we add flow to the situation, exactly the same thing applies: if the dynamic pressure at the connection to the tank drops, the air will expand to restore -- again, very nearly -- the previous pressure. If the dynamic pressure at the tank rises, the air will compress. Therefore, to a first approximation, the dynamic pressure at the point at which the tank is connected will not change

As an aside, it is very desirable that the absolute dynamic pressure at the inlet to a centrifugal pump be controlled so as to not drop below a certain pressure, depending on the characteristics of the pump and the temperature and characteristics of the fluid in question. The most reliable way to do this is to ensure that the static pressure is adequate -- and then to place an expansion tank at or very near that point to ensure that the absolute dynamic pressure doesn't change. Hence pumping away.

Now if we consider the case with two expansion tanks, we again have to realise that the function of the expansion tanks is to allow the system contents to expand or contract, as the case may be, with temperature. In the static situation, they will respond equally. However, in the dynamic situation, the absolute pressure at the two tanks will differ, depending on the flow and losses between them. One of them, therefore, will compress more than the other, but the total change in volume of the two tanks will be the same as if they were just one tank of larger size. Indeed, if there is no temperature change, there will be no change in the total volume. Therefore, if there is flow, since the absolute pressure at the upstream one will be greater than the absolute pressure at the downstream one, the upstream one must compress and the downstream one expand -- and somewhere in between there will be no absolute pressure change. The location of this point will vary depending on the configuration of the piping or channels or whatever connecting them (and, incidentally, the flow rate, since head loss is not linearly related to flow).

It might be of interest to point out that the above consideration of two expansion tanks is strictly valid only for constant flow conditions. If the flow is varying, the momentum of the moving fluid must be considered. In a heating system, this is trivial. In a larger system, however -- say a public water supply, or a hydroelectric power plant -- it is anything but trivial and it may be necessary to have considerable "expansion" capability in the system -- usually in the form of standpipes or even damping basins or reservoirs -- to avoid catastrophic water hammer effects with rapid changes in flow.

I guess in summary, one has to realise that an expansion tank has an influence only with regard to temperature change (or other volume change), and that that influence is to maintain the absolute pressure in the system at the expansion tank constant.

hot_rod

It is possible to run a closed hydronic system without an expansion tank. The GEO guys do it all the time
I've done some small bathrooms with a 2 or 6 gallon electric HW tank without expansion. Pressure change is small enough from 65- 100° to not pop a relief. Some suggest the pex wall will take the expansion also.
That being said, really no reason not to add an expansion tank. I have also used the small tankless DHW expansion tanks for small hydronic systems. You'll see those small globe shaped exp tanks on demo at trade shows also

EdTheHeaterMan

I had a feeling that Jamie would be able to explain it in the simplest most concise way.

Thanks, Jamie.

Thing is ... I understand everything he said, Just don't ask me to splane it to yous guys That is Jamies Job

Warmwater

Thanks for the comments guys!

DanHolohan

Gil Carlson wrote about having two compression tanks hooked up to different points. He said that the PONPC would establish somewhere in between the two tanks. This would cause one of the tanks to lose water and the other tank to gain water (he was writing about plain-steel tanks, not diaphragm tanks).

I've seen this happen when a cast-iron radiator filled with air and wasn't notices. Often, the radiator was in a stairwell. It pretended to be a second compression tank and it cause the real tank in the boiler room to constantly lose its air cushion. We bled the radiator and that solved the problem.

Zman

DanHolohan said:

Gil Carlson wrote about having two compression tanks hooked up to different points. He said that the PONPC would establish somewhere in between the two tanks. This would cause one of the tanks to lose water and the other tank to gain water (he was writing about plain-steel tanks, not diaphragm tanks).

I've seen this happen when a cast-iron radiator filled with air and wasn't notices. Often, the radiator was in a stairwell. It pretended to be a second compression tank and it cause the real tank in the boiler room to constantly lose its air cushion. We bled the radiator and that solved the problem.

A property I consult for has 2 diaphragm tanks that were inadvertently installed on opposite ends of a 9,000,000 Btu, 2,000 gallon snowmelt system. When I noticed the condition, I eagerly isolated the tanks to see what the effect was. As Dan said, they average out (more or less).

Warmwater

Thanks guys. Interesting topic.