alpha pump flow rate not enough for some rads

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • DBMandrake
    Automated Home Legend
    • Sep 2014
    • 2361

    #16
    Originally posted by MJNewton View Post
    Note that you can't reliably read the temperature of a metallic (reflective) surface with an IR sensor as its emissivity will be way short of (and lacking consistency with) the near-perfect body that is assumed. Put a small square of black insulation tape on first as this has an emissivity of ~0.95 which is likely what the gun has been calibrated for.
    Your point about reflective metallic surfaces is correct, however most modern radiators are painted and therefore don't have reflective metallic surfaces... Paint makes a good black body radiator and the kind used on radiators has an emissivity fairly close to 0.95. For any painted radiator there is no need to add tape, and tape will slow down the response time of the reading as tape after all is a fairly good insulator. You don't really want the added thermal lag of the tape when you're trying to quickly compare the surface temperatures of many radiators as they are heating up.

    The only time you might need to do something like this would be an unpainted antique cast iron radiator - but even then an oxidised surface still has a very high emissivity compared to polished metal.
    Also make sure you hold the sensor right up to the surface - they essentially average across an imaginary 'cone' of IR input and hence the further away you are the more background radiation you'll pick up (lowering the apparent temperature).
    Good point which I should have made - hold the gun about an inch away from the panel to ensure that its field of view doesn't pick up parts of the wall etc or the reading will be artificially low.

    Comment

    • DBMandrake
      Automated Home Legend
      • Sep 2014
      • 2361

      #17
      Originally posted by therealfronty View Post
      Hmmm this is interesting, new user here and just seen this thread, I also have a Grundfos Alpha2 pump like this, also a Viessmann 26kw boiler, except mine is y-plan open vented, and I have 13 radiators. I also have struggled with the pump settings, for years I have fiddled with the different settings and found that the middle fixed speed setting "II" (pump shows around 20w consumption) is the only one that provides a satisfactory amount of heat with an acceptable level of noise.

      I just recently had another play, reading the instructions it suggests that auto-adapt or PP2 are the best settings for a system like mine, and I would love for the pump to auto-adjust based on how many TRVs are open, but with both settings the pump only draws between 5-8w and the boiler will shut down after a few minutes because the flow temperature overshoots the set point, you can see the flow temp slowly drop as the pump continues to run until eventually the boiler fires and overheats the water again and so on - I end up waking up to a cold house in the mornings. I think the water simply isn't moving fast enough through the hex. I tried CP2 as well with similar results, it's just not running fast enough. On fixed speed 2 everything runs fine, the boiler stays lit and the flow temp stabilises around 68-69 degrees.

      It's a shame there isn't a way to adjust these PP settings, because I have this fancy pump with all this "intelligence" built in but end up ignoring all that and just running it like a standard fixed speed pump. There is a slight noise from some of the radiator return valves but it's bearable, what makes more noise are the motors in the HR92, the one in our bedroom wakes me up, but that's another topic!
      Yes this is exactly what happened to me when I tried to use the variable speed modes on my pump - all three variable modes were completely useless on my system because the pump would slow right down to a trickle even with all radiators open, let alone with only a few open.

      There is too much flow resistance in my system (consisting of about a 3 metre run of a 28mm manifold which then branches off to long runs of mostly 8mm microbore) for the load sensing pump to operate properly. When all my radiators are open it thinks most of them are closed and throttles right down! Needless to say my boiler isn't happy.

      So I use it in either slow or medium fixed speed modes. Slow is enough for all but the depths of winter and saves a bit of power while I use Medium in the winter to squeeze that last bit of performance out.

      Comment

      • sandyman
        Automated Home Sr Member
        • Mar 2016
        • 85

        #18
        Originally posted by therealfronty View Post
        Hmmm this is interesting, new user here and just seen this thread, I also have a Grundfos Alpha2 pump like this, also a Viessmann 26kw boiler, except mine is y-plan open vented, and I have 13 radiators. I also have struggled with the pump settings, for years I have fiddled with the different settings and found that the middle fixed speed setting "II" (pump shows around 20w consumption) is the only one that provides a satisfactory amount of heat with an acceptable level of noise.

        I just recently had another play, reading the instructions it suggests that auto-adapt or PP2 are the best settings for a system like mine, and I would love for the pump to auto-adjust based on how many TRVs are open, but with both settings the pump only draws between 5-8w and the boiler will shut down after a few minutes because the flow temperature overshoots the set point, you can see the flow temp slowly drop as the pump continues to run until eventually the boiler fires and overheats the water again and so on - I end up waking up to a cold house in the mornings. I think the water simply isn't moving fast enough through the hex. I tried CP2 as well with similar results, it's just not running fast enough. On fixed speed 2 everything runs fine, the boiler stays lit and the flow temp stabilises around 68-69 degrees.

        It's a shame there isn't a way to adjust these PP settings, because I have this fancy pump with all this "intelligence" built in but end up ignoring all that and just running it like a standard fixed speed pump. There is a slight noise from some of the radiator return valves but it's bearable, what makes more noise are the motors in the HR92, the one in our bedroom wakes me up, but that's another topic!
        I've been playing with my settings after replacing my bypass valve, which was bust. with the bypass set to a high value, pump in PP3 mode, warm up times are really slow, pump stays at low speed, boiler takes a long time to warm up. similar to you. whereas III is much quicker and gets the job done. Interestingly if I drop the bypass pressure, this allows some short-circuiting , the warm up time in PP3 mode improves. For these tests I am only running a single heat destination , my DHW cylinder, so the resistance of the rest of the circuit is a constant.
        what I notice happening is that the pump speed in PP3 mode seems to be proportional to water temp as well. this fits with some of the grundfos documentation. With the partial short circuit of a bypass open a bit, the return temp to the boiler goes up more quickly , so the boiler has to increase flow temp to keep the delta-T correct so burns a bit harder, then the pump speeds up some more, then its a positive feedback loop until we end up with a properly hot flow circuit at the target temp, boiler modulating to keep it there, and the pump running quite fast (still in PP3 mode). It does appear to me that cold-start is the problem - assuming you have an ABV, maybe you could try a bit of short-circuit?
        what would be ideal would be for them to make a pump with a "boost mode" relay (that you could trigger on a timer or some IOT cleverness) or even better a linkage to evohome. so based on a large call for heat, cold system in the morning or whatever, the system would set the pump to mode III to get things warmed up as fast as poss, and then for tickover it could run it on Auto/PP/CP mode saving on noise and circulating pump power.

        In my case, I will persevere with balancing and check my TRV locations as others have suggested, but I suspect I will need to keep to a PP mode for noise reasons. I am considering splitting the system with a buffer, will allow introduction of ASHP as well, in which case it will allow a fixed speed on the primary side and a variable on the secondary.

        Comment

        • MJNewton
          Automated Home Jr Member
          • Jan 2021
          • 21

          #19
          Originally posted by DBMandrake View Post
          Your point about reflective metallic surfaces is correct, however most modern radiators are painted and therefore don't have reflective metallic surfaces... Paint makes a good black body radiator and the kind used on radiators has an emissivity fairly close to 0.95. For any painted radiator there is no need to add tape, and tape will slow down the response time of the reading as tape after all is a fairly good insulator. You don't really want the added thermal lag of the tape when you're trying to quickly compare the surface temperatures of many radiators as they are heating up.
          It is usual practice to measure the pipes or valve tails, not the radiator itself, in order to ensure consistency. It is after all the flow and return temperatures you are trying to measure. Paint is almost certainly going to have variable emissivity depending on colour, pigment and compound. A piece of tape won't introduce any thermal lag whatsoever (it is too thin; indeed thinner than paint - but neither will make any difference anyway as the lag is already in the order of minutes due to the thermal mass of the metal component parts) but will significantly improve accuracy. Black PVC tape is very consistent in emissivity hence why it is the go-to surface for accurate IR temperature readings - most non-contact thermometer instructions will specify its use if they pass any comment at all.

          There are numerous lookup tables for emissivity figures - such as this one - where'll you see black insulating tape always ranked right up at the top. Interestingly this particular table includes something called 3M Black Velvet Coating 9560 with an emissivity of 1.00 and whilst unnecessary and likely an expensive specialist product you would likely need to adjust your thermometer to use that accurately as it will almost not have been calibrated to assume such a perfect black body surface.
          Last edited by MJNewton; 16 May 2021, 12:29 AM.

          Comment

          • MJNewton
            Automated Home Jr Member
            • Jan 2021
            • 21

            #20
            Originally posted by sandyman View Post
            With the partial short circuit of a bypass open a bit, the return temp to the boiler goes up more quickly , so the boiler has to increase flow temp to keep the delta-T correct so burns a bit harder
            Was this through observation? I would expect the opposite. The boiler should be trying to maintain the set target temperature and whether it's tracking the flow, return or both if the return temperature is increasing it should be modulating down to try and meet the target. That's one of the purposes of the bypass - it is effectively signalling to the boiler (in a rather crude, but nevertheless effective, way) that the boiler's heat supply is exceeding the demand (or sink capacity, if lower) and so needs to reduce the heat output (eventually to the point of cycling if it can't modulate low enough). This is why if the bypass is opening too soon (wrongly set, stuck open etc) then heat-up times will be extended and overall output lowered as the boiler is always having to hold itself back.

            I suspect the reason you saw the heat-up time improve with a partially opened bypass was down to the increased flow this allowed through the boiler hence greater dissipation of heat being produced - even if it was just around whatever length of pipework you have between the boiler and bypass because, if I understand your test setup correctly, this will contribute a relatively high additional flow path when compared to just the coil in your cylinder. You would be better off including some radiators in any such tests as it'll balance out the heat source:sink ratio and minimise the impact of what should otherwise be subtle influencing factors.
            Last edited by MJNewton; 16 May 2021, 12:27 AM.

            Comment

            • DBMandrake
              Automated Home Legend
              • Sep 2014
              • 2361

              #21
              Originally posted by MJNewton View Post
              It is usual practice to measure the pipes or valve tails, not the radiator itself, in order to ensure consistency. It is after all the flow and return temperatures you are trying to measure.
              "Usual practice" perhaps, but that assumes you are trying to adjust the balancing of the radiators to produce a certain temperature differential across each radiator. This is indeed the traditional approach but as I mentioned in my original post after having tried this method in the past I would argue strongly that this approach is fundamentally flawed regardless of being "usual practice", (it's trying to optimise for the wrong parameter) and I mention a couple of possible aggravating factors that contribute to this, such as short radiator vs long radiator where it's difficult to even achieve a large temperature drop on a short radiator without excessively restricting the flow.

              IMHO balancing via lockshield valves is about equalising the flow resistance to the different radiator branches so they all get a fair chance when all radiators a full on during initial warm up, to prevent any radiators from being left behind or completely starved. And for this purpose measuring the flow/return differential is meaningless, what you're interested in is the average temperature of each radiator (hence measuring half way along to get a quick single point measurement of the average of flow and return temperatures) and you want to be able to do it very quickly, as you need to measure all radiators quickly multiple times during a 10 minute warm up period to gauge the relative warm up rate of each radiator compared to the others.

              Clamp sensors are just too slow to get clamped onto pipes, take measurements then move onto the next radiator - while an IR gun isn't as precise its good enough and extremely quick, and if you follow the right measurement technique, very repeatable. Clamp sensors also don't work well if the pipes are painted as the paint will slow down the response of the sensor when the sensor is applied.

              At the end of this initial balancing you can then try to more accurately measure the differential across each radiator and if the differential is too small or too large on most radiators, increase or decrease the pump speed a bit, and in fact the overall temperature drop of the radiators should be set with the pump speed not the lockshield valves.
              Paint is almost certainly going to have variable emissivity depending on colour, pigment and compound. A piece of tape won't introduce any thermal lag whatsoever (it is too thin; indeed thinner than paint - but neither will make any difference anyway as the lag is already in the order of minutes due to the thermal mass of the metal component parts) but will significantly improve accuracy. Black PVC tape is very consistent in emissivity hence why it is the go-to surface for accurate IR temperature readings - most non-contact thermometer instructions will specify its use if they pass any comment at all.
              Unless the radiators are very different from one another (modern painted radiators vs old cast iron ones etc) then none of this is necessary. You don't need a super accurate absolute temperature reading, in the method I've described you're trying to get a comparison of how quickly the different radiators heat up - and for this relative readings are all you need. The emissivity of the IR gun doesn't even need to be calibrated as you don't care what the exact figures are, just how one radiator compares to another.

              If you have radiators with vastly different surface finish then sure, put a big square of black tape on the middle upper portion of each radiator, I have no problem with that. Just be sure that the entire surface area of the tape is pressed onto the metal and that half of it isn't sitting out in the air, and make sure it's big enough to fully cover the viewing cone of the sensor from the distance it is held at. You also have to be careful not to hold it too close for too long as heating up the plastic inside the IR gun sight can cause inaccuracy.
              Last edited by DBMandrake; 17 May 2021, 09:23 AM.

              Comment

              Working...
              X