Viessmann Vitodens 100-w trouble maintaining central heating temp

[Gasmk1]

I can't answer that but what I can say is if you look at my first ever post answering you there a picture of how it should be done ? , a small low loss header with a separate pump fitted would also do the job on your existing set up look at the Viessmann installation guide it may give you a better understanding . Kop

Totally agree

 

[John.g]

If the pump speed is limited to 72% then the 7.5M pump becomes < 4M which probably isn't enough for 12mm piping but strange that the system worked OK apparently for years.

 

[dr911t]

I can't answer that but what I can say is if you look at my first ever post answering you there a picture of how it should be done ? , a small low loss header with a separate pump fitted would also do the job on your existing set up look at the Viessmann installation guide it may give you a better understanding . Kop

@king of pipes - wholeheartedly agree. After all the research and help from this forum I know much better of what a proper UFH should look like.
The OEM circulation pump inside the boiler is the Grundfos UPM3 15-75 which also comes as a UFH pump. If I were to add another pump to the system, wouldn't it basically be running two pumps? Is it because of the thickness of pipes that so much pumping pressure would be needed? If we had say 10 rads in the flat instead, would an external pump not be necessary then because of larger pipes? We are on a one level 120m2 with 16mm PERT-AL pipes.

Is that the system with the manifold circ pump and where you run the boiler at say 60/65C and UFH at 35/35C which reduces the primary flow rate substantially?. Your post #23.

@John.g - are you referring to @king of pipes ' setup?

If the pump speed is limited to 72% then the 7.5M pump becomes < 4M which probably isn't enough for 12mm piping but strange that the system worked OK apparently for years.

Piping is 16mm. 120m2.
Agree it's strange that it worked but I'm not so sure the manifold area is the problem since the noise behind the pump and possible circulation seem to be an issue also when just DHW is being used (if I turn off central heating).

 

[John.g]

Well, I would be quite surprised if one pump couldn't give adequate flow through 16mm piping, my daughter has a mixture of rads/UFH serviced with a 16 year old Grundfos selectric 6M (one) pump, all heating working perfectly.

Don't know what length the loops are but assuming 100M then the pressure drop through each loop is 0.152M @ 1LPM, 0.54M @ 2LPM and 0.82M at 2.5LPM, these numbers become 0.18M, 0.66M & 1.0M for 120M loops.

I know you are in a dilemma after spending all that money but if I were you (and should have suggested this previously) I would go away and buy a 0 to 10M pressure gauge, install it in the UFH flow side of the manifold and run your system, that will tell a lot and may save you a pile of money at the end of the day. Also find out if that circ pump is restricted to 72% speed.

Yes, I was referring to KOPs post/photo.

 

[dr911t]

Well, I would be quite surprised if one pump couldn't give adequate flow through 16mm piping, my daughter has a mixture of rads/UFH serviced with a 16 year old Grundfos selectric 6M (one) pump, all heating working perfectly.

Don't know what length the loops are but assuming 100M then the pressure drop through each loop is 0.152M @ 1LPM, 0.54M @ 2LPM and 0.82M at 2.5LPM, these numbers become 0.18M, 0.66M & 1.0M for 120M loops.

I know you are in a dilemma after spending all that money but if I were you (and should have suggested this previously) I would go away and buy a 0 to 10M pressure gauge, install it in the UFH flow side of the manifold and run your system, that will tell a lot and may save you a pile of money at the end of the day. Also find out if that circ pump is restricted to 72% speed.

Yes, I was referring to KOPs post/photo.

Thanks.. I would assume also loops are less than 100M since some of the larger 20-40m2 rooms were split into 2 or 3 loops but I don't have the schematics used during construction. Circulation pump is at 80% speed.

 

[John.g]

Didn't use proper wall thickness for that 16mm piping so I think a ID of 12mm so the corrected numbers for 100M loops are 0.32M @ 1LPM, 1.16M @ 2LPM & 1.75M @ 2.5LPM so the boiler circ pump running at 80% should still give a flow rate > 2LPM.

 

[Ric2013]

I've seen far worse installations. In this case no mixing valve is required since the boiler is itself supplying the water at the required low temperature. Short pipe runs to the manifold mean not too much loss of head. If it used to work, then the basic design may not be at fault.

If the OP isn't getting more than 1 lpm even with only one circuit open, there is very obviously something wrong. The fact that water is flowing at all would seem to rule out airlocks in the UFH pipework, but there may be some kind of algal blockage that has developed over the summer season. Nothing to stop the OP isolating the UFH manifold from the boiler and mains flushing a circuit using a hose to see what the water is like and if there is any blockage.

EDIT If number of circuits open does not appear to affect the flow through the circuits that are open then pump would appear to be on a constant pressure setting.

 

[John.g]

Have you ever noticed/noted the manifold pressure (if PG installed) on a UFH system without a manifold pump.

 

[Ric2013]

Have you ever noticed/noted the manifold pressure (if PG installed) on a UFH system without a manifold pump.

If that's a question for me, then afraid the answer is no, but it would be very useful information if someone here has it.

Thinking about it, it's not uncommon to see a 2kW radiator run on a 10m or more run of 15mm copper tube which has similar internal bore to 16mm MLCP UFH pipe and this could be achieved at a fairly low head. A UFH circuit may be much longer, but won't be 2kW. So there oughtn't be a lack of pressure to drive at least one circuit (...unless the pump ports/boiler H Ex etc. are obstructed) or the UFH circuits are themselves obstructed.

If an inline flow gauge is an option, what if the OP were to create a short circuit by attaching a braided hose across the two hose connectors at the end of manifold? If a decent flow were then obtained, that would point to the issue being in the UFH circuits themselves rather than in the boiler itself?

 

[John.g]

His low flow rate may be partly be because the "Maximum speed setting in the delivered condition" for the boiler is 80% so this makes it a 4.5M pump, don't know if this setting can just be changed to the 100% rating of 6.8M which the previous pump may have been set to?. It also mentions residual head, not sure if this means the head available after allowing for the Hx. Setting may be changeable from the boiler menu??.

 

[Ric2013]

Hi John,

To probably slightly mis-quote "The Lion the Witch and the Wardrobe", the OP states the system worked correctly for 4 years and so either he is lying, or he is telling the truth, or he is mad (or mistaken).

I think we need to work on the assumption that the OP's statement that the system has worked for 4 years is correct as otherwise this is a waste of everyone's time. I don't see why we are trying to prove the technical feasibility of a no external pump setup when 4 years of working order has already done this. My own UFH manifold at my own home tees off the primary flow in 15mm which is wrong by every rule of thumb and I had to defend the decision on technical grounds when discussing with colleagues until after I'd installed it, after which, the fact that it functioned perfectly made such defence of my 'undersized' pipework unnecessary.

While it may be interesting to explain why the OP's system should work, if it cannot be made to work, then some of the others may be right to say such a single-pump system can't work, but that implies that the OP is not correct to say it formerly did work. We cannot all be correct.

Looking at your chart, we can have up to 5.5lpm total before we fall below a 4 metre head, and if the circuits will have enough pressure on a 3m head (minus resistance of primaries and boiler itself) to run properly, then we have 13lpm available flow which should give each circuit about 1.5lpm which ought to be enough really. The problem is we don't know if a 3m head is available at the circuits, but an undersized pump would be indicated by dropping off of flow to each individual circuit as more and more circuits are opened. Looking at the data for the UPM3(K) AUTO XX-50 130/180 (N) which is, I think, the kind of pump that might be fitted to a UFH system, its highest constant pressure setting is only 3m anyway, so unless the boiler/primaries in the OP's case are robbing more than 1.5m head, he should have at least enough oomph to run 2 circuits at 1.5lpm with the pump you have shown the curves for.

At this point, we need to wait to hear back from the OP. From that funny noise in the boiler, I'm thinking the UFH circuits are not at fault and something is up with the boiler (which was the original assumption) but it would be nice to rule out the UFH, and it might be something really silly like someone had fiddled with the flow adjusters on the manifold, or an isolator valve is remaining partly shut although it looks open.

What about this kind of approach?:

How might the OP prove the manifold is at fault? I think your idea of testing the flow at a specific working head is good, but may be a job for a plumber rather than a DIY one, though I suppose we could guide the OP through the process.

How might the OP prove the boiler is at fault? It would be very easy to fit a low loss header and then find, if the boiler is at fault, this has 'cured' the flow issue through the manifold, but things still wouldn't work properly and it would probably reduce the boiler efficiency as well as be an expensive experiment.

 

[John.g]

The cheapest way to see what the actual head is, is to install a PG which the OP I would think is capable of doing since he flushed out the manifold, the other approach is to ask Viessmann if the circ pump can be set to 100% because I don't think anyone would be happy to run a UFH with a single 4.5M pump which this pump effectively is at 80% speed. The original pump may well have been set to 100% which would go a long way to explaining normal performance for a few years.

 

[Ric2013]

He could quite easily install a pressure gauge on the 3/4" BSP hose connector isolators. It would presumably show differential pressure relative to atmospheric so he'd need to subtract the system charge pressure from the pressure created by the pump in action, but it would give an idea.

Does the following sound right? If the assumed blockage is between the return rail and the pump inlet on the return side, you'd see a good pressure measured at the flow rail and at the return rail. If the blockage is the UFH circuits themselves (which might just be acting normally due to their inherent resistance), you'd see a lower pressure on the return rail than on the flow rail, and if there is not a decent pressure on either rail then the blockage is between the pump outlet and the flow rail, or the pump is not running properly for some reason.

If that's your plan, I think it's an excellent way of locating the problem and will help even if the problem proves to be internal to the boiler.

 

[John.g]

Yes, you are correct, he would really need 2 PGs, one on the flow, the other on the return, for some reason or other I thought it was a OV system so the PGs would need to be 0 to 3bar which makes it difficult to get a accurate differential reading of say 0.2bar between the two but worth trying at any rate?.

 

[dr911t]

OP here. Thanks for all the idea brainstorm.

I've isolated the manifold and flushed every loop for 5' with tap mains. Water is clear. No blockage. Bit of air in some loops but not much. Flow rates way over 4LPM (max of the gauges). I don't think there is a blockage in the manifold itself because as previously mentioned, even with one circuit open, flow rate is max 1.5LPM. If the boiler was fine but the manifold or UFH was the issue then hot water would be running perfect when CH is off however the boiler runs hot... 75-85C at times as it modulates up and down.

Viessmann & the tech's current theory is blocked main HX (calcified). Pump cannot get enough pressure through the main heat exchanger to move the water around. The sound possibly has something to do with it. Pump is brand new (few weeks old). This theory might be plausible although not sure if it explains the sound from the hydroblock. However Viessmann tech has done a 2h power wash with acid (don't ask me which) to remove limescale in the HX, didn't change anything in terms of boiler behavior or noise. There's nothing else behind there besides the PRV and automatic air vent. Could it be some back-flow into the pump due to the alleged 'blockage' in the HX?

Another strange thing, we do not have an outside temperature sensor and weather dependent mode on (installer couldn't get access to drill on a proper outdoor place or some reason like that) and the manual says without wether dependent mode, pump speed can't be changed, however I seem to be able to still do it. I can change the speed up to 100%. At 100% it just makes more noise like a car engine on high revs, but LPM flow doesn't change!

I'm going to look into finding some pressure gauges and fit them at the fill/drain connections of the manifold.

@Ric2013 I might qualify for the mad/mistaken part but if my memory serves me right, this noise never existed in previous years, boiler was running at low temps in CH without issues at 35-45C flow and flow meters were showing more than 2LPM so something went bonkers this year. All flow meters are fully open during this troubleshooting.

 

[John.g]

Why should the main Hx be calcified, its a closed system and except there is a constant leak somewhere and the system is being topped up with a auto fill PRV. If the system is manually topped once/twice/annum and the filling valve(s) are then shut then calcified Hx should not be a problem. Was this Hx ever changed?. Can understand if it was the DHW Hx was getting calcified due to constant mains water through one side of it.

One simple test you can do is to carefully note the boiler pressure before fire up and circ pump start up, as soon as the pump starts up, again carefully note the boiler pressure, if the E.vessel is teed in before the pump suction then the pressure should rise by the pump head, if teed in after the pump then the pressure may fall by the pump head.

Another manifold test which may be worth doing is to disconnect the flow and return, get one 0 to 1bar PG and install it in the flow side of the manifold or install it in the mains filling hose, then open the mains filling hose isolating valve slowly and note the pressure with flow through all the loops on together at say 1LPM/loop, 1.5LPM/loop, 2LPM/loop & 2.5LPM/loop. This should give a indication of the pump (differential) head required at these flow rates.

 

[king of pipes]

To many if buts and maybes going on here all the info the op needs is in the manufacturers instructions it's just a matter of studying them closely and putting it into practice, the boiler needs to be installed and configured to supply the correct amount of flow through the heat exchanger in my opinion it isn't .

 

[John.g]

Can you send on a link please KOP to that manual or just the pressure loss through the HX.

The OP states that he has seen flow rates of 2LPM= 18LPM/9loops or a boiler flow rate of 1080 LPH. My calculations show a required head of 1.16M (@2LPM) assuming 100M/loop, allow another 25% for elbows etc then the required head is ~ 1.5M or allowing another 72% still results in a required head of "only" 2M.
You can tell from your manual, the pressure loss through the (clean) Hx. I don't know what this is but assuming that the total pump head required is 4.5M then in your opinion should the boiler supply this without any problem or not?, ie a flowrate of 1080LPH @ 4.5M or so?.

dr might inform of the output of his boiler, if not already done so.

 

[king of pipes]

Here you go John busy on breakdowns I don't think it will I could be wrong thinking it needs a low loss header ? Kop

 

[dr911t]

Why should the main Hx be calcified, its a closed system and except there is a constant leak somewhere and the system is being topped up with a auto fill PRV. If the system is manually topped once/twice/annum and the filling valve(s) are then shut then calcified Hx should not be a problem. Was this Hx ever changed?. Can understand if it was the DHW Hx was getting calcified due to constant mains water through one side of it.

Closed system. Annual top-ups. No previous flush of heating water. Unknown what original installer put in the water if anything. The UFH is 120m2 so with an average of 5m of pipe per m2 and 0.15-0.17L per m of piping I'm estimating around 90-100L of water in the UFH. Is that enough to calcify a main HX? It's the original one. Boiler was new and installed 5 years ago.
The plate heat exchanger was recently replaced. Looking back I don't think it needed to, a flush and some chemicals would have been enough. It didn't look too calcified.

dr might inform of the output of his boiler, if not already done so.

It's the 2017 100-W B1KA 24kw combi.

it needs a low loss header

I was under the impression these are more for commercial setups with multiple boilers providing separation from multiple circulating pumps. I think if they would have done the job properly or had I had more knowledge at the time, then an external circulating pump, mixing thermostatic valve, low loss header, pressure/temp gauges, a filter, all should have been fitted.

 

[John.g]

Lack of inhibitor may cause Hx blockage due to corrosion.
Still can't find the dP across the main (clean) Hx.


Did you note the change in boiler pressure on circ pump start up.?

 

[hometech]

You have a return temperature back to the boiler that is not a true reflection of the system temp.

Hot water flow is returning back to the boiler too early without dissipating heat which means the flow is mixing with the return and raising the temperature of the return higher than it would otherwise be.

This will always be a function of the pipework route. If the route has not been changed then a valve is causing the issue.

Buy a digital thermometer and measure the temperatures along the pipe, at valves, joints etc. Where the temp rises or drops within a few centimeters is your problem point. it only takes a very small rise in return to create this issue.

As you are using low temperature water for circulation, your problem will be unimaginable to 99% of plumbers who only know about basic combi boilers and room thermostats set at 22 degrees.

If you want a more scientific approach rather than the random guesses, have a look at heat geek's site. Go back to basics, its so refreshing

Renew the heating industry - HeatGeek

Heat Geek offers expert advice on all aspects of the heating industry. Browse our knowledge hub online now.

www.heatgeek.com

 

[dr911t]

Lack of inhibitor may cause Hx blockage due to corrosion.
Still can't find the dP across the main (clean) Hx.


Did you note the change in boiler pressure on circ pump start up.?

CH On - Pump 70% - 1.5Bar - Flow ~1L/min
CH On - Pump 100% - 1.6Bar - Flow ~1L/min - pump noise increases. - boiler temp drops by about 4-5C after a few minutes.
Boiler Off - 1.4Bar
DHW On - CH off - 1.5Bar after pump start

The pressure meter does seem to wiggle a bit between 1.5-1.6 Bar during CH operation... bit odd.

In regards to adding a pressure gauge to the manifold: found one that has a 1/8 male connection, problem is the manifold intakes are 1/2 male connection too so having some trouble finding an adapter. I'd rather not attempt to remove the auto air vents. I

 

[dr911t]

You have a return temperature back to the boiler that is not a true reflection of the system temp.

Hot water flow is returning back to the boiler too early without dissipating heat which means the flow is mixing with the return and raising the temperature of the return higher than it would otherwise be.

This will always be a function of the pipework route. If the route has not been changed then a valve is causing the issue.

Buy a digital thermometer and measure the temperatures along the pipe, at valves, joints etc. Where the temp rises or drops within a few centimeters is your problem point. it only takes a very small rise in return to create this issue.

As you are using low temperature water for circulation, your problem will be unimaginable to 99% of plumbers who only know about basic combi boilers and room thermostats set at 22 degrees.

If you want a more scientific approach rather than the random guesses, have a look at heat geek's site. Go back to basics, its so refreshing

Renew the heating industry - HeatGeek

Heat Geek offers expert advice on all aspects of the heating industry. Browse our knowledge hub online now.

www.heatgeek.com

I think the problem is the opposite. Too much temperature drop due to not enough flow.

Heating Rotary Knob Setting: Between 4 & 5
Boiler temp: 49-50C
Heating flow: 47C
Heating return: 27.5C
Manifold flow meters: ~1L/min flow meters open fully - all thermostats on, 9 loops.

I have it set like this (for now to have heating until the issue is fixed) because it only operates on these temps and higher, any setting below just causes it to short cycle.

 

[IND_Nick]

As already discussed it definitely appears to be a circulation issues, I don't deal with Viessmann but are there any gauze filters on the return side of the boiler. Many manufacturers have them in the return valve or close by. If missed during a flush it can stay blocked for a while and force flow through the plate to plate rather than the desired heating circuit. Just a thought, may not be relevant on the Viessann though.