The flow and return temperature variation now with the system going full pelt trying to re-heat the house is 30 degrees - return = 44 flow 74 degrees at the boiler. Ovbiously something is very wrong? Maybe the pump is broken - or the Magnaflow I fitted has enormous flow resistance?
Maths to see if it's right - not guesswork.
- Thread starter Useroftheforce
- Start date
Sponsored Links
Why have you shut all the LSV's?
The bypass has to be adjusted with all rad valves fully open and any TRV heads removed. This will give the minimum flow resistance round the circuit.
The pump speed is adjusted to give an 11C differential at the input to the ABV. The ABV is then adjusted to give a 9C differential at the boiler. When the system is in operation and radiator TRV's close down the flow rate will reduce. As the pump runs at a constant speed, the pump head will increase, so the pressure differential across the ABV will increase and the valve will open further to maintain the flow through the boiler.
The boiler High thermostat should be kicking in at about 82C. If nothing happens until it reaches 96C, the chances are the High stat has failed and the overheat stat is kicking in. What happens if you run the boiler on Low? The temperature should be about 65C.
I can't see where you got you figures from. A 320W radiator will require a flow rate of 0.32/(4.18 x 11) = 0.00696 kg/sec not 0.007022 which you gave. You also used a flow rate of 0.06 when finding the flow resistance in Table 2. This gave a figure of 0.018 which gives 0.018 x 37.44 = 0.67392 metres. As the table does not have data for flow rates lower than that 0.010 fro a 15mm pipe you will have to use that. This gives a resistance of 0.001m per metre, i.e 37.44 x 0.001 =0.03744 metres.
The calculations will give you two figures: the circuit resistance, in metres; and the flow rate in kg/sec. The "metres" is actually metres of water, so this can be easily converted to kPa. If we assume that the density of water does not change with temperature, the calculated flow rate is also the same figure in litres/second. This can then be easily converted to litres/minute of cubic metres per hour.
Calculating the index circuit can be very complicated as you have to take into account the complete system. The section headed Tabulation Method is a good guide to what to do. I created a spreadsheet to do mine, and then there is a lot of guesswork as I can't determine the exact pipe layout. Don't forget to include the boiler resistance in the calculation. For the 30/50Si it's 0.71m.
Thanks for the reply - I think my calculations were for a 323w rad. But we arrived at almost the same figures... I was in a bit of a hurry - but your example has made it clearer.
I closed the LSV's to check the system balance. I could have just closed all the TRV's for what I was aiming to acheive. That was. Increase the flow to maximum through the primary circuit to bring the flow and output temperature as close together as possible. That involved eliminating the heat sinks.
I can see that I need to spend time in the spreadsheet and also with a tape measure to work things out. Although it is becoming pretty obvious as I suspected, that there is a design fault in my system that needs correcting.
I closed the LSV's to check the system balance. I could have just closed all the TRV's for what I was aiming to acheive. That was. Increase the flow to maximum through the primary circuit to bring the flow and output temperature as close together as possible. That involved eliminating the heat sinks.
I can see that I need to spend time in the spreadsheet and also with a tape measure to work things out. Although it is becoming pretty obvious as I suspected, that there is a design fault in my system that needs correcting.
What you are simulating with all LSVs closed is the case where all TRVs have closed, so the flow rate will be at a minimum and the pump delivering the highest pressure. The ABV should therefore be open in these circumstances to make sure the flow rate through the boiler is maintained. If the valve is fully open and you cannot achieve the correct temperature gradient the pump may be undersized.
You haven't commented on my suggestion that the High thermostat may be u/s.
Each turn (assuming soldered not compression) adds 0.6m so you have an equivalent pipe length of 14 + (10 x 0.6) =20m. At an assumed temperature drop of 9C across the boiler the flow rate will be 14.8 / (4.18 x 9) = 0.393 kg/s ≈ 0.4 kg/s. From Table 2, at 0.4kg/s the flow resistance is 0.079m/metre. So the flow resistance through 20m of pipe is 20 x 0.079 = 1.58m. If we then add the boiler resistance of 0.71m we get a resistance of 2.29m; and this is before we get to the radiator circuit. The 15/50 pump can only supply a head of 3m at 0.4 l/s, which means there is only 3-2.29 = 0.71m left for the radiator circuit and that does not take into account any loss through motorized valves.
It looks as if transam was right when he said, in a very early reply:
transam said:
Sponsored Links
"It's worse than that Jim"
The actual calculated length for the pipes to and from the pump when I worked it all out properly with a tape measure was 19.9m @ 22mm.
There are 2 tees right out of the boiler - a short stub ontop for bleeding. The main flow, therefore does a sharp right angle, through the tee, in and out of the boiler.
Then there is the magnaclean on the return side to consider +2 elbows.
There are a total of 15 elbows + 1 more T (into the bypass circuit) in that circuit including the bypass. the ABV, which after all said and done is not a piece of straight pipe, I have assumed is an elbow.
That works out at 30.4m total @ 9 degrees drop as you say.
30/50si is a 50000 BTU boiler = 14.65 KW. 14.65 / (4.18 x 9) = 0.389L/s or kg/s. Therefore = 0.079 resistance at that speed for 22mm. So 30.4 X 0.079 = 2.416h + boiler at .71 = 3.116 m. So effectively I am out... I am not even sure that with this kind of resistance that a 15/60 will have the required grunt to power the index and other radiators as well. Possible major rethink on the way - but I shall calculate first.
I did note that you believe that the high stat may be broken - but it was only replaced about 3 months ago with a brand new one. Since the flow through the boiler has now been proved to be so inadequate I would imagine that is not helping the situation at all. I expect it's easy for the burner to overshoot the temperature since the lag between the hot water and the thermostat being heated by same is too large.
I am running the system on the lower setting at the moment and it's much happier there.
The actual calculated length for the pipes to and from the pump when I worked it all out properly with a tape measure was 19.9m @ 22mm.
There are 2 tees right out of the boiler - a short stub ontop for bleeding. The main flow, therefore does a sharp right angle, through the tee, in and out of the boiler.
Then there is the magnaclean on the return side to consider +2 elbows.
There are a total of 15 elbows + 1 more T (into the bypass circuit) in that circuit including the bypass. the ABV, which after all said and done is not a piece of straight pipe, I have assumed is an elbow.
That works out at 30.4m total @ 9 degrees drop as you say.
30/50si is a 50000 BTU boiler = 14.65 KW. 14.65 / (4.18 x 9) = 0.389L/s or kg/s. Therefore = 0.079 resistance at that speed for 22mm. So 30.4 X 0.079 = 2.416h + boiler at .71 = 3.116 m. So effectively I am out... I am not even sure that with this kind of resistance that a 15/60 will have the required grunt to power the index and other radiators as well. Possible major rethink on the way - but I shall calculate first.
I did note that you believe that the high stat may be broken - but it was only replaced about 3 months ago with a brand new one. Since the flow through the boiler has now been proved to be so inadequate I would imagine that is not helping the situation at all. I expect it's easy for the burner to overshoot the temperature since the lag between the hot water and the thermostat being heated by same is too large.
I am running the system on the lower setting at the moment and it's much happier there.
Re checked cacls the actual pipe distance should have been 34.9m - therefore Head is actually 3.153 M.
I have completed all the calculations now. Unfortunately the total resistance in the index circuit is 6.28m. This might be a tad over what it really is, but not allot I am afraid.
I did notice that in the run right from the pump to the 1st Tee, the heat mass index required was 0.55 which is more than 22mm will supply. In reality though the emitters require in total 46 / 17 = .369l/s. But of course the system design is so awful that the resistance increases that wildly.
There isn't an easy way to reduce this resistance as far as I can see. Would it be possible to add another pump in series in the primary circuit? Two 15/30 pumps would be able to provide 6m head at the no2 setting @ 4l per second. And just about stay within the 2.5m maximum pipe velocity.
??
I did notice that in the run right from the pump to the 1st Tee, the heat mass index required was 0.55 which is more than 22mm will supply. In reality though the emitters require in total 46 / 17 = .369l/s. But of course the system design is so awful that the resistance increases that wildly.
There isn't an easy way to reduce this resistance as far as I can see. Would it be possible to add another pump in series in the primary circuit? Two 15/30 pumps would be able to provide 6m head at the no2 setting @ 4l per second. And just about stay within the 2.5m maximum pipe velocity.
??
I really am out of my depth here - but I have a proposal...
I could T into the feed and return on the primary circuit in the cellar. With an ABV on a short pipe between the two proposed T's. Where I am thinking of doing that the head would = 2.29m .4l/s inc boiler. So, if I stuck another 15/50 pump in just before that ABV on the flow side then I am thinking that would take care of the boiler flow required.
That would then leave the existing pump on the first floor with nearly all of it's capacity to provide the head to the radiators which have a requirement of 2.99m @ .4l/s.
It's hard to visualize if this is utter rubbish without being a fluid engineer
I could T into the feed and return on the primary circuit in the cellar. With an ABV on a short pipe between the two proposed T's. Where I am thinking of doing that the head would = 2.29m .4l/s inc boiler. So, if I stuck another 15/50 pump in just before that ABV on the flow side then I am thinking that would take care of the boiler flow required.
That would then leave the existing pump on the first floor with nearly all of it's capacity to provide the head to the radiators which have a requirement of 2.99m @ .4l/s.
It's hard to visualize if this is utter rubbish without being a fluid engineer
It appears I was wrong - this is a familiar feeling now
The actual Head in the index circuit is 4.55m @ 11 degrees drop. The reason why I got it wrong was forgetting that the full section mass flow only travels in the 22mm pipe and not the 15mm spur to each emitter.
Now it's correct, it calculates the mass for the rad against the short 15mm spurs and fittings. Then also calculates the sum of the mass flow for the whole section at each T off, againsts the 22mm pipe size fitted + fittings.
Even so 4.55m is still alot at 4l/s. (The UPS 25-80 looks like it would do it, if I need to get one large pump.)
I have wrestled all day to get excel to do a lookup against the mass flow index table and have finally sussed it.
In case anyone else decides they need to perform this hideous operation the formula is below.
=INDEX('Lookup Table'!$A$6:$C$28,MATCH(C27,'Lookup Table'!$A$6:$A$28,-1),2)
Explanation::
Input the table two from "Copper Tubes in Domestic Heating Systems" onto an excel tab called "Lookup Table". This is the important bit - upside down. Start with the large numbers at the top and work down the page to the smaller numbers in the table.
The formula does the following: Takes all the values found at 'Lookup Table'!$A$6:$C$28 The $ means the reference won't change when you copy the cell. So in effect; find from these range of cells A6:C28. MATCH means take this value and find it = C27 in my case. This contains the mass flow index that you are interested in. The bit after the next , says find the mass flow index in the range A6:A28 on the lookup table. This is where you have put the Flow Rate in Kg/s numbers. The next thing is important. -1 says where no exact match is found return the next highest number. So in other words give me a slightly higher resistance. The final figure is which column to return the relative value from. So 2 means in my case, return the value from the second column. This is where I placed the 15mm pipe resistances. 3 would mean from the 22mm pipe column and so on for other pipe sizes and resistances.
What this allows me to do is vary the temperature drop across all the emitters and see the effect on the total head of the system all by changing one number. This is possible because the mass flow index is calculated using the formula (rad watts/1000)/(B8*B5) Where B8 = 4.18 and B5 = the required temperature drop.
All this is leading up to saying because the house only needs 11kw and the emitters deliver 17kw I could run the rads with a 16 degree drop and they will still deliver enough heat. I can then from my calculations see that if I set the ABV to 1.24m it will provide the required pressure in the rad circuit, leaving the rest to head back to the boiler and keep it running at the correct flow rate and temperature drop.
This will reduce the system head to 4.01m which is only just outside the capacity of a 30/60 unit @ 4ls with no other changes required.
So transam - you were right!
Anyone got any advice on pumping options? Two pumps in series? Or one large one?
The actual Head in the index circuit is 4.55m @ 11 degrees drop. The reason why I got it wrong was forgetting that the full section mass flow only travels in the 22mm pipe and not the 15mm spur to each emitter.
Now it's correct, it calculates the mass for the rad against the short 15mm spurs and fittings. Then also calculates the sum of the mass flow for the whole section at each T off, againsts the 22mm pipe size fitted + fittings.
Even so 4.55m is still alot at 4l/s. (The UPS 25-80 looks like it would do it, if I need to get one large pump.)
I have wrestled all day to get excel to do a lookup against the mass flow index table and have finally sussed it.
In case anyone else decides they need to perform this hideous operation the formula is below.
=INDEX('Lookup Table'!$A$6:$C$28,MATCH(C27,'Lookup Table'!$A$6:$A$28,-1),2)
Explanation::
Input the table two from "Copper Tubes in Domestic Heating Systems" onto an excel tab called "Lookup Table". This is the important bit - upside down. Start with the large numbers at the top and work down the page to the smaller numbers in the table.
The formula does the following: Takes all the values found at 'Lookup Table'!$A$6:$C$28 The $ means the reference won't change when you copy the cell. So in effect; find from these range of cells A6:C28. MATCH means take this value and find it = C27 in my case. This contains the mass flow index that you are interested in. The bit after the next , says find the mass flow index in the range A6:A28 on the lookup table. This is where you have put the Flow Rate in Kg/s numbers. The next thing is important. -1 says where no exact match is found return the next highest number. So in other words give me a slightly higher resistance. The final figure is which column to return the relative value from. So 2 means in my case, return the value from the second column. This is where I placed the 15mm pipe resistances. 3 would mean from the 22mm pipe column and so on for other pipe sizes and resistances.
What this allows me to do is vary the temperature drop across all the emitters and see the effect on the total head of the system all by changing one number. This is possible because the mass flow index is calculated using the formula (rad watts/1000)/(B8*B5) Where B8 = 4.18 and B5 = the required temperature drop.
All this is leading up to saying because the house only needs 11kw and the emitters deliver 17kw I could run the rads with a 16 degree drop and they will still deliver enough heat. I can then from my calculations see that if I set the ABV to 1.24m it will provide the required pressure in the rad circuit, leaving the rest to head back to the boiler and keep it running at the correct flow rate and temperature drop.
This will reduce the system head to 4.01m which is only just outside the capacity of a 30/60 unit @ 4ls with no other changes required.
So transam - you were right!
Anyone got any advice on pumping options? Two pumps in series? Or one large one?
have read through this complete post and the calcs and knowledge being used are all very admiral, but, (theres always one).
Get the heat exchanger removed and have it descaled, automotive radiator repairers are good for this. they use very strong chemicals and can pressure test it afterwards. Its low water content copper.
This boiler was designed donkeys years ago, well before auto by-passes where considered on domestic systems. Have the manual primary by-pass installed to maintain correct flow rate and temp difference across boiler, you might have to install a balance valve on system primary return to achieve desired result. Fit a 15/60 pump.
Quite a number of older boilers Radiation, Thorns, Glowworms required this arrangement.
Get the heat exchanger removed and have it descaled, automotive radiator repairers are good for this. they use very strong chemicals and can pressure test it afterwards. Its low water content copper.
This boiler was designed donkeys years ago, well before auto by-passes where considered on domestic systems. Have the manual primary by-pass installed to maintain correct flow rate and temp difference across boiler, you might have to install a balance valve on system primary return to achieve desired result. Fit a 15/60 pump.
Quite a number of older boilers Radiation, Thorns, Glowworms required this arrangement.
It's a steep learning curve.It appears I was wrong - this is a familiar feeling now
Are you sure it is 4 litres per second? That would be true if you had a 150kW boiler! I think you mean 0.4 litres per sec. Assuming your Head calculation is correct, a 15/60 pump would be too small and the 25/80 will be to large. According to the Grundfos pump sizer, the ideal pump is the Magna 25/60. But they are expensive pumps - over £400. Alternatives are the UPS 32/55 (£250) and UPS 40/50 (£220).
My formula is:
INDEX(Resistance, MATCH(K13,Flow,-1), MATCH(G13,Pipe,))
I have used name ranges: "Flow" is the first column, "Pipe" is the pipe size row and "Resistance" is the body of the table.
I also took into account the variations due to temperature difference, but have gone a step further than you. I can vary all temps: flow, return, room and outside. It then calculates the heat loss for each room and the radiator multiplier. From this it calculates the required radiator output, the flow rates though the rads and the heat loss in the pipes. At standard temps my head is 5.46M and I have a 15/50 pump!
Like you I have a long run of pipe from the boiler (kitchen) to pump (airing cupboard); it's about 21M total equivalent length. If that wasn't there the head would only be about 2.5M. I'm not worried as, when a new boiler is installed working at a 20C differential the total head reduces to 2.3M.
One solution to your problem is to install a "low loss header" and a second pump. Look it up!
One thing the designers of the system did not take into consideration was heat loss in the pipes; I have calculated this as 4.86KW. Fortunately most of this is useful heat, pipes between floors or behind plasterboard. The rads total 13kW so in theory I need an 18kW boiler. Fortunately the house only needs 7.5kW to heat it so, depending on the contribution from the pipes, the rads may only need to produce as little as 2.5kW.
This probably explains why, even in the recent cold weather when the outside temperature was sometimes -8C, the house maintains a temperature of 21C with the rad temps about 60C or less.
Thanks heatingman. I drained the system today and re-installed a permanent 15mm bypass loop with a gate valve on it. In fact I just bridged the existing 22mm bypass circuit with the ABV in it. ASCII art coming up...
---- Something like that where the primary flow is not shown to the left of
| | the diagram. The left line is the 15mm with gate valve. The right is
---- The 22mm (22mm top and bottom) with the ABV in the bottom right
Corner.
I guess by doing that (once I get the 15-60) I can set the correct 9 degree drop and the have the ABV operate as well as all the TRV's close. At the moment drop across the boiler has halved to 15 degrees. I am still therefore only flowing 0.232 l/s across the burner.
I'll clean the heat exchanger if needed after new pump fitted.
D_hailsham. While I had it all apart I was checking the stats. It appears the boiler is fitted with two hi stat's and a limit stat. They were connected incorrectly (christ knows who installed and serviced this system before!) so the high stetting was running off the limit stat!! Anyway - I have rectified that. Must remember to order a new low stat.
The upshot of correcting as much as possible the boiler flow is that there isn't sufficient pressure in the system as we pretty much know. So off to get a new pump tomorrow.
I was thinking further. Since the radiators are oversized and the house only needs 11kw heat, perhaps I could run the radiators with a larger temperature drop.. say 20 degrees drop. The would further reduce the required head.
I had a look at that calculation you posted from (the danfos website I think) but formulas aren't really my thing.
Would you mind walking me though it please?
Just to save typing on your part the t1 tr and t2 are all obvious. I can see that we divide whatever the result of the top part is by t1-t2. I can also see that we divide the differential differences by each other in the ()section. t1-tr / t2-tr. This is all pretty easy I guess. So where we have 80/60/20. That means we have 60/40 = 1.5. Multiply 49.33 X ln (1.33?? what is this ln business, I can see what n is but not ln.) 49.33x1.33 = 65.609 - now what? then there's another n on the outside of the brackets - oh lord... have you got an excel formula for it?
I am very, very grateful for your help solving this problem.
---- Something like that where the primary flow is not shown to the left of
| | the diagram. The left line is the 15mm with gate valve. The right is
---- The 22mm (22mm top and bottom) with the ABV in the bottom right
Corner.
I guess by doing that (once I get the 15-60) I can set the correct 9 degree drop and the have the ABV operate as well as all the TRV's close. At the moment drop across the boiler has halved to 15 degrees. I am still therefore only flowing 0.232 l/s across the burner.
I'll clean the heat exchanger if needed after new pump fitted.
D_hailsham. While I had it all apart I was checking the stats. It appears the boiler is fitted with two hi stat's and a limit stat. They were connected incorrectly (christ knows who installed and serviced this system before!) so the high stetting was running off the limit stat!! Anyway - I have rectified that. Must remember to order a new low stat.
The upshot of correcting as much as possible the boiler flow is that there isn't sufficient pressure in the system as we pretty much know. So off to get a new pump tomorrow.
I was thinking further. Since the radiators are oversized and the house only needs 11kw heat, perhaps I could run the radiators with a larger temperature drop.. say 20 degrees drop. The would further reduce the required head.
I had a look at that calculation you posted from (the danfos website I think) but formulas aren't really my thing.
Would you mind walking me though it please?
Just to save typing on your part the t1 tr and t2 are all obvious. I can see that we divide whatever the result of the top part is by t1-t2. I can also see that we divide the differential differences by each other in the ()section. t1-tr / t2-tr. This is all pretty easy I guess. So where we have 80/60/20. That means we have 60/40 = 1.5. Multiply 49.33 X ln (1.33?? what is this ln business, I can see what n is but not ln.) 49.33x1.33 = 65.609 - now what? then there's another n on the outside of the brackets - oh lord... have you got an excel formula for it?
I am very, very grateful for your help solving this problem.
As you have found out, you can't draw pics easily. That's because, in their infinite wisdom, the programmers of this site have decided to automatically reduce all "unwanted" blank spaces to one space. The only sensible solution is to create the drawing in Paint and post is as an image
Yes, the formula is in the Danfoss Heating Book (very useful but much is irrelevant to UK domestic systems), but I also found it elsewhere on the net.
"ln" means the natural logarithm, i.e log to base e. It has nothing to do with the "n", which is usually taken as 1.33.
The Excel formula is: (49.83*LN((D2-G2)/(E2-G2))/(D2-E2))^1.33
where
D2 = flow temp
E2 = return temp
G2 = room temp
You divide the nominal (catalogue) output of the rad by the calculated factor to get the actual output.
If you want to play around with the formula go to Heat emissions from radiators and heating panels. You can enter the numbers and see how it changes.
The only problem is that is assumes a flow temp of 80C and return of 60C. That's why the constant is 49.32. This means that rad outputs are slightly lower than they would be by about 2%.
I have installed the 15-60 and refitted the bypass with a gate valve and removed the ABV. In fact I just re-made the entire bypass loop out of 22mm, coz the old one was a bit of a dog end.
Where I am at now is the following.
All the rads are set to give a 22 degree drop. The house is lovely and warm, although it takes a while to get there. The boiler doesn't short cycle at all all. It just burns and burns. There is no kettling noise from the boiler.
Heating wise, at the moment it's only been run with all the TRV heads off for balancing and testing.
However I am still not really sure things are as they should be.
Before I started the balancing, as a test, I set the boiler to run on the low stat and fired it up with the HW circuit only. The HW tank is right by the Y valve. Then I opened the 22mm gate valve on the bypass all the way and shut the balance gate valve on the HW circuit. This meant, primary only at 22mm X 32.7m @0.4ls which the new pump easilly, should do. However I could not bring the boiler down to less than 11 degrees drop. The head in that circuit is 3.3m 0.4ls so the new pump should be able to bring the temperature closer than 9 degrees with no other load.
Then I started considering the heat loss in the pipes on the primary circuit. according to this link http://www.engineeringtoolbox.com/copper-pipe-heat-loss-d_19.html the loss would be about 1.2kw for an uninsulated 20m run 22mm at 55 degrees. Would that affect the temperature?
I also thought that it could be as has been said that the HE is full of crud as has already been said. Since the boiler has been run for most of it's life with reduced flow and we are in a heavy chalk area... ??
Despite this it was 2 degrees last night so I set the gate valve to a turn open as per the boiler manual and balanced the system at 22 degrees drop. This setting provides 1.1kw more than the loss of the house. This gave 15 - 16 degrees drop at the boiler.
If I open the gate valve more (attempting required 9 degree drop), there isn't enough pressure in the system to feed the radiators. If I open the rads to try and get 11 degrees, same problem, not enough pressure.
I guess therefore - either something is blocked. There isn't the heat from the boiler there should be, or my head calcs are way out.
The system is virtually silent in opperation.
Where I am at now is the following.
All the rads are set to give a 22 degree drop. The house is lovely and warm, although it takes a while to get there. The boiler doesn't short cycle at all all. It just burns and burns. There is no kettling noise from the boiler.
Heating wise, at the moment it's only been run with all the TRV heads off for balancing and testing.
However I am still not really sure things are as they should be.
Before I started the balancing, as a test, I set the boiler to run on the low stat and fired it up with the HW circuit only. The HW tank is right by the Y valve. Then I opened the 22mm gate valve on the bypass all the way and shut the balance gate valve on the HW circuit. This meant, primary only at 22mm X 32.7m @0.4ls which the new pump easilly, should do. However I could not bring the boiler down to less than 11 degrees drop. The head in that circuit is 3.3m 0.4ls so the new pump should be able to bring the temperature closer than 9 degrees with no other load.
Then I started considering the heat loss in the pipes on the primary circuit. according to this link http://www.engineeringtoolbox.com/copper-pipe-heat-loss-d_19.html the loss would be about 1.2kw for an uninsulated 20m run 22mm at 55 degrees. Would that affect the temperature?
I also thought that it could be as has been said that the HE is full of crud as has already been said. Since the boiler has been run for most of it's life with reduced flow and we are in a heavy chalk area... ??
Despite this it was 2 degrees last night so I set the gate valve to a turn open as per the boiler manual and balanced the system at 22 degrees drop. This setting provides 1.1kw more than the loss of the house. This gave 15 - 16 degrees drop at the boiler.
If I open the gate valve more (attempting required 9 degree drop), there isn't enough pressure in the system to feed the radiators. If I open the rads to try and get 11 degrees, same problem, not enough pressure.
I guess therefore - either something is blocked. There isn't the heat from the boiler there should be, or my head calcs are way out.
The system is virtually silent in opperation.
Do you mean that you have balanced the rads to give a 22C drop?
Hang on a sec! When you did this test, was the motorized valve open to allow a flow through the radiator circuit? If it was, your circuit was not simply the 22mm from boiler to bypass and back again. It would include all the heating circuit as well, so the head would not be 3.3m.
Presumably you did the test with the pump on speed 3. What happens if you use speed 2? What temperature difference do you get when you do exactly the same test?
Of course! If there was no heat loss, the flow and return temps would be the same!
As I said before, the 15/60 is not up to the job.
DIYnot Local
Staff member
If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.
Select the supplier or trade you require, enter your location to begin your search.
Are you a trade or supplier? You can create your listing free at DIYnot Local
Sponsored Links
