System design advice please!

[Biggles..]

This boiler needs big primaries and big pumps.

Boiler resistance is the issue,if it can't shift it's heat quick enough when comparing f and r temps,as soon ad above 20deg it'll go into a wait status,flow will cool,return won't rise boiler fires again,cycle continues.

Pipework and pump are guaranteed your problem.

Ben you have to remember 15/60 only good to about 30kw also this boiler has a head of effectively 4metres without the rest of the system(.index circuit).

 

[bengasman]

...Ben you have to remember 15/60 only good to about 30kw also this boiler has a head of effectively 4metres without the rest of the system(.index circuit).

Almost half a bar resistance for the boiler? That hex must look like a tea strainer.

I know that the old open vent systems needed a lot of pump capacity. Oddly enough, the only pump problem I come across on systems I install (rarely more than 20 kw dissipation on the rads) is that I can't squeeze the lockshields enough to realise 20 degree drop without stopping the flow completely. WB CDI's have water channels big enough to run pebbles through, hardly any resistance.

 

[D_Hailsham]

With the boiler in the cellar and a boiler with a very high internal resistance, it sound an ideal candidate for a low loss header.

I don't know why it shows the primary pump in the return. It also does not show the secondary pump(s).

 

[ajrobb]

I may be barking up the wrong tree here, as you say opening the bypass reduces boiler cycling, but have you balanced the radiators? Also, how is it with the water heating off (just thinking that could need restricting too)?

 

[Onetap]

I think you could be assuming that the problem is only due to the concealed pipework being under-sized and/or the pump being inadequate. There could also be other issues involved; air locks, debris blocking the pipes, concealed valves (either tiled over or under access covers that haven't yet been found), unbalanced system, etc.. I would expect the house, and top floor, to be uniformly but inadequately heated if the problem was solely due to low flow, and that in turn due to pipes & pump too small. I would also have expected to get some flow to the top floor, even if only by shutting off other radiators.

The pump sounds to be under-sized so a LLH is what I'd probably do, but you need to tell the owner at the start that there could be other, undetected faults and you can't guarantee satisfaction, not having a crystal ball. He may get the hump with you if he spends money and doesn't get what he'd expected. I'd use a variable speed pump on the secondary, since you'll have to take a guesstimate at the size and a bigger fixed speed pump could introduce new problems with TRV noise.

The LLH may increase the return temperature, if secondary flow < primary flow rate, but that is lesser problem compared to getting adequate flow to all the rads.

You can also get reduced secondary flow temperature, if secondary flow>primary flow, due to some reversed flow in the LLH.

I'd think about supplying the cylinder from a pumped secondary circuit, leaving the primary flow constant, rather than have it changing when a zone valve opened.

Is it practical to contact the previous owners to ask about the problem or get the details of the installer?

 

[spraggo]

Is it practical to contact the previous owners to ask about the problem or get the details of the installer?

I doubt whether you will get much luck with the installer, I would imagine that the indians got to him first.

 

[ajrobb]

Interested consumer speaking. I might need a similar sized boiler to replace a pair of Keston Celcius 25 boilers. I don't want to lose condensing efficiency by keeping the existing arrangement with separate pumps for boiler flow and radiator flow.

Just looked at the manual. Blimey, a 4m water gauge pressure drop across the boiler sounds a lot (what's that, about 0.42 bar?). The way they have the ABV in figure 4.2 in the manual, you have to add this boiler pressure drop to the ABV setting to get the working head for the pump. So, for an ABV factory setting of 0.2 bar, that is about 0.62 bar at 1633 l/h (boiler flow). If the pipes are too small, you'll have to crank up the ABV pressure (to get all that flow through the radiators) and, consequently, the pump pressure too.

Incidentally, I calculate that 1633 l/h at 38 kW is about 20°C temperature rise across the boiler.

 

[D_Hailsham]

Just looked at the manual. Blimey, a 4m water gauge pressure drop across the boiler sounds a lot (what's that, about 0.42 bar?). The way they have the ABV in figure 4.2 in the manual, you have to add this boiler pressure drop to the ABV setting to get the working head for the pump. So, for an ABV factory setting of 0.2 bar, that is about 0.62 bar at 1633 l/h (boiler flow). If the pipes are too small, you'll have to crank up the ABV pressure (to get all that flow through the radiators) and, consequently, the pump pressure.

It doesn't work like that.

Let's assume that the pump has been properly sized to deliver the required flow rate against the total head of the system (boiler+rads), which is calculated assuming that all rads are open. When TRV closes the flow rate reduces and, from the pump graph, the head delivered by the pump increases.

All the ABV has to do is negate the increase in the head so the flow returns to the required value.

Incidentally, I calculate that 1633 l/h at 38 kW is about 20°C temperature rise across the boiler.

That's correct.

The advantage of a wide differential is that the flow rate is approximately half that of the traditional 11°C differential, so the pump can be smaller as well as the pipes. The only disadvantage is that the output of a radiator is reduced by about 20%.

If you think a 4m boiler resistance is high, it would be off the scale if an 11°C differential was used.

Thought

Could this be the cause of the OP's problem, i.e the system is trying to run with an 11°C differential?

 

[ajrobb]

It doesn't work like that.

I appreciate that IF the system was OK, the pressure drop across the radiators with TRVs open should be less than the ABV setting.

With the TRVs open and the radiators lockshield valves balanced to give 20°C drop across each radiator, the pressure drop across the boiler should be at LEAST 4.05m water gauge to get at least 1633 l/h flow. The pump speed should be adjusted so that at least one radiator has to have its lockshield fully open.

If the pressure drop across the boiler is less than 4.05m, then the radiators are too small. The pump rate is too low and should be increased (accept the reduced temperature differential). The boiler will then be able to operate at maximum power until the radiators warm up.

If the pressure drop across the boiler is greater than 4.05m with a 20°C drop across the radiators, then the radiators are too big. Drop the pump speed to bring the pressure across the boiler down to 4.05m. The temperature drop will try to increase above 20°C. The feed temperature won't reach set point until some of the TRVs close but the boiler can run efficiently at maximum output while the house is warming up.

The ABV shouldn't begin to open until several TRVs are closed.

 

[D_Hailsham]

The pump speed should be adjusted so that at least one radiator has to have its lockshield fully open.

Where do you get that idea from?

If the pressure drop across the boiler is less than 4.05m, then the radiators are too small.

Not necessarily. It could mean that the boiler is oversized.

The pump rate is too low and should be increased (accept the reduced temperature differential).

You set the pump to give the correct flow rate, as determined by the temperature drop across the boiler.

The boiler will then be able to operate at maximum power until the radiators warm up.

But you have to match the output of the boiler to that of the radiators. It all comes down to calculating the heat loss from the house and then choosing the correct components to meet that.

If the pressure drop across the boiler is greater than 4.05m with a 20°C drop across the radiators, then the radiators are too big.

How did you work that out?

 

[ajrobb]

My answers are based on the existing boiler and radiators:

The pump speed should be adjusted so that at least one radiator has to have its lockshield fully open.

Where do you get that idea from?

'Cos it's daft to have the pump pushing hard against closed lockshields.

If the pressure drop across the boiler is less than 4.05m, then the radiators are too small.

Not necessarily. It could mean that the boiler is oversized.

Alright, too small for the boiler then.

The pump rate is too low and should be increased (accept the reduced temperature differential).

You set the pump to give the correct flow rate, as determined by the temperature drop across the boiler.

Flow rate and temperature drop are not the same thing. The flow rate is a function of pressure drop.

The boiler will then be able to operate at maximum power until the radiators warm up.

But you have to match the output of the boiler to that of the radiators. It all comes down to calculating the heat loss from the house and then choosing the correct components to meet that.

In an ideal world... but the OP is working with existing boiler and rads.

If the pressure drop across the boiler is greater than 4.05m with a 20°C drop across the radiators, then the radiators are too big.

How did you work that out?

Obviously, I mean too big for the boiler. The flow is too great to match the boiler.

 

[D_Hailsham]

My answers are based on the existing boiler and radiators:

'Cos it's daft to have the pump pushing hard against closed lockshields.

But you don't close a lockshield, you set it to give the correct flow through the radiator. In any case a lockshield may appear to provide a very small pathway for the water when it is almost closed, compared to the size of a 15mm pipe, but it is perfectly adequate for the flow rate.

You set the pump speed to provide the required temperature drop at the boiler and across the rads. A properly balanced system will often run with a lower pump speed than an unbalanced one.

Flow rate and temperature drop are not the same thing. The flow rate is a function of pressure drop.

I did not say flow rate and temperature differential are the same thing. I said that you can indirectly measure the flow rate by measuring the temperature differential. I agree that flow rate is a function of pressure differential, but how they are related will depend on the size and design of the radiator. The formula used for calculating the flow rate through a boiler, given the output and temperature differential is also valid for a radiator.

The boiler will then be able to operate at maximum power until the radiators warm up.

But you have to match the output of the boiler to that of the radiators. It all comes down to calculating the heat loss from the house and then choosing the correct components to meet that.

In an ideal world... but the OP is working with existing boiler and rads.

I agree it was how it should be done and the OP has to live with what he has. But you seem to be suggesting that, if the OP has a 38kW boiler and only 20kW of rads, that's OK as the boiler will be able to give out all 38kW while the rads are heating up. I think you will find that the water temperature will rise so quickly that the boiler will immediately modulate down to provide the required output.

 

[ajrobb]

You set the pump speed to provide the required temperature drop at the boiler and across the rads. A properly balanced system will often run with a lower pump speed than an unbalanced one.

We don't know if the boiler size is right for the radiators. If it is, then the pump speed should be set to give 1633 l/h and balancing the radiators at 20°C will work fine. If the radiators are 'too big' then that limits the feed temperature that can be achieved (effectively de-rating the radiators to match the boiler and increasing condensing efficiency to boot). If the radiators are too small to dissipate the full boiler power there is the hot water heating to consider; this can absorb a fair bit.

It might take 10 to 15 minutes for 30 radiators to reach operating temperature. In that time, the boiler can operate at full chat even if the radiators are too small, but only if the flow rate is high enough. Once the boiler is cycling on the room 'stat, that could account for much of the time it is on.

Basically, if the flow through the boiler never reaches the design rate (1633 l/h in this case), the boiler cannot operate properly at full power.

 

[Andygasman2010]

It's already a sealed system (installer hasn't even bothered connecting up the system and cylinder PRVs, they're just left internal)

I did think about possible restrictions in pipework elsewhere but if there are any they're under the floor i.e. inaccessible. I'm quoting to upgrade what I can see.

I did warn the guy that the improvements i'm quoting are genuine and will improve things but there may still be other issues. I'm pretty sure we'll get the top floor going for him as it doesn't work at all at the minute.

The radiators nearest the boiler, ground floor hall and lounge big doubles, are really hot yet for 10 mins or so after firing up the boiler cycles like I described.

I think the undersized pump and primaries can't deal with 38kw of heat and that needs sorting and the rads need properly balancing. Once that's done I personally think we'll get away without touching the rest under his tiled floors! Not ideal but you have to play the cards you're delt

 

[Andygasman2010]

I may be barking up the wrong tree here, as you say opening the bypass reduces boiler cycling, but have you balanced the radiators? Also, how is it with the water heating off (just thinking that could need restricting too)?

I've not touched the rads other than feel them for heat, balancing will probably help but it'd only be masking the real problems so i'll do it after i've upgraded the pump/primary pipework.

The cylinder is next to the boiler/plan so i've been trying it with the hot water off for now. That's in 22mm so it works fine, it'll need a balancing valve fitting at some point.

I assume when I open the bypass and it works better, it's because i'm giving the pump a lower path of resistance?? If I do this it gets up to temp quicker.