PDA

View Full Version : Flow rate through low loss header



TiredGeek
26-04-2011, 04:14 PM
OK, for various reasons I have fitted a low loss header into the system to help with flow through my 2 x 14kw Mitsubishi EcoDan ASHP.

The header is a 42mm pipe with 28mm pipes in and out (4 each side) and there are two water pumps drawing from it and feeding the ecodans (1 to each) on the return side. The radiator side of the header has 4 water pumps, 2 on flow and 2 on return. All the pumps are the same spec: Grundfos 15/60 240vac. The two feeding the Ecodans are set to fixed full speed, the rad side are on variable speed 2.

The thing is, I'm only getting an indicated flow of 25L/min on the flow setters that are in the ASHP side of the circuit. The reason this header was added was to increase the flow of water through the ASHP, but it's pretty much the same as before it was fitted when the water was just coming back from the rad circuit.

I'm no expert but I thought I'd be seeing much higher flow rates through the header than this as there is now no restriction from the rads.

What do you guys think?

Brian_UK
26-04-2011, 11:51 PM
So the ASHP pumps are also 15/60's ?

What is the resistance through the circuit and how does that fall on the pump curve?
http://www.plumbnation.co.uk/site/grundfos-ups-15-50-selectric-130-bare-pump/Grundfos%20UPS%20Selectric%20Pumps%20Brochure.pdf

mad fridgie
27-04-2011, 12:28 AM
Your low less header is to give you a consistent flow to your heat pumps, regardless of the flow through the radiators. If your indicators are those little red things in a tube, then a little change in indication can actually mean a big change in actual flow (my opinion only) I do not find them to be that great.
Is the system working well?

MikeHolm
27-04-2011, 12:42 AM
Unless you have a lot of other restrictions on each line, that pump has lots of head capacity for the current flow but it depends on how much more flow you want (35-40L/m?). In Canada I would be using 1.25" copper or I guess that is 32mm there. Perhaps you are near the piping limits. What elbows, valves and check valves, flowsetter restrictions are there?

Bigfreeze
27-04-2011, 01:23 PM
Unless you have a lot of other restrictions on each line, that pump has lots of head capacity for the current flow but it depends on how much more flow you want (35-40L/m?). In Canada I would be using 1.25" copper or I guess that is 32mm there. Perhaps you are near the piping limits. What elbows, valves and check valves, flowsetter restrictions are there?

I agree, 1" pipe for a 14kw unit especially at this time of year is quite small as that unit is probably producing between 16 - 17kw at the moment, depending on how its controlled

TiredGeek
27-04-2011, 01:58 PM
Hi guys, thanks for the replies :)

Pumps: sorry, wasn't clear enough. All 6 water pumps are Grundfos Alpha 2L 15/60. http://www.plumbnation.co.uk/site/grundfos-alpha2-l-15-60-domestic-circulator-bare-pump/grundfos-alpha2-l-installation-guide.pdf
There are 4 on the radiator side of the header, they are all set to PP2 (page 13).
The two on the ASHP side are set to constant curve, speed 3.

Piping: All is 28mm, isolator valves are full bore, 2 way valves are 28mm, swept curves used where possible, minimal use of elbows. It was designed with flow in mind :)

The restrictions come in the form of the flow setters (one for each ASHP) and the flexi pipes that come with the units. Mitsubishi spec a 28mm pipe size, but then supply flow setters with what appears to be a 15mm bore and about 20mm bore in the flex pipes. To me that is where the restriction is, but I'm no plumber ;)
Here is an installation instruction sheet that shows both the flexi pipes and flow setters just like I have, page 14, http://www.iceenergy.co.uk/documentation/manuals/Ecodan_FTC2_Manual.pdf

That installation sheet is pretty similar to the one I had, probably made up by Mitsubishi and modded by the supplying companies......

Brian_UK, sorry, I don't know the resistance, but like I say, it was designed with as few sharp elbows and other restrictions as possible.
Between the header and the ASHP, all in there is: 1 x 15/60 pump, 4 x 28mm elbows, ~4 metres of 28mm copper, 2 x ~300mm long flexi pipes & 1 flow setter. That's everything, flow and returns.

I was thinking that the flow would have been in the 30+ range. Mitsubishi reckon 18L/min is the bare minimum, ~40L/min is preferable, and that is what I was hoping to get near.
I notice on the sheet I linked to that it says 25-40 L/min is where it should be, I didn't have that info when mine were first plumbed in.

Hope that gives you more info to work with :)

Bigfreeze
27-04-2011, 04:21 PM
I can't see you getting much more than 25L a minute through the 15mm flow setters to be honest. Why not take them out and judge you flow by the water temps across your condenser

MikeHolm
28-04-2011, 12:30 AM
Absolutely, as BF says, get rid of as much restriction as you can. I think Mits is nuts to use the size of flexi hose as they do. Depending on the type of hose, the restriction could be up to 4 times that of copper per metre length and the connectors don't help. In any piping length, controls such as valves can be only 1 size smaller than the needed pipe diameter without having a very marked effect on the total resistance to flow.

As I see it, the piping needs to be 32mm, the valves must be full bore and the flow setters should be sized up or the pump must make up for the head loss. Also, make sure the flow setter has 6" or more of straight pipe fore and aft to give a proper reading. Hope this helps.

Jon Glanfield
28-04-2011, 03:21 PM
Hi Tired Geek, we have managed to achieve 25 lpm as you did without low loss headers using 2 15-60 pumps and 28mm pipework, as others have said though Mitsubishi's design criteria flies in the face of others e.g. Daikin who would require 35mm on their twin fan units.

The flow setter is a restriction and others are available (not sure if I can say where) that we use on the 35mm circs for other heat pumps, likewise the flexis are too, and solid connections at the same bore size are not detrimental provided pipework is adequately supported.

From various posts on here though be wary of the panacea that LLH are seen as, we have used them and small buffers and seen a multitude of issues in terms of flow rate, spiking return temps and mixing resulting in cycling.

A decent sized buffer makes much more sense with commensurately sized flow and returns to the heat pump. However as people have pointed out before it is rare to see 2 Ecodans banked unless they are linked via a Step Controller or on separate heating/hot water systems within 1 property.

We have had 1 that iced up as yours have done, but this was due to the system volume being too small, an issue that we flagged with the suppliers and Mitsubishi, both of whom disregarded our concerns, both are now visiting to investigate further, this is not an isolated case either from what we can deduce. Mitsi have never been able to confirm to us what minimum system volume they require, but instead quote min and flow rates as the design requirement.

If loops close or TRVs shut the system can be left with a very small volume of heat energy to perform defrost, all of which has to get through the restriction of an auto bypass which compromises flow rate further.

I was also advised recently that if the heat demand from the property were to cease part way through an Ecodan defrost cycle, it does not complete the cycle leaving excess frost on the fins. I have no empirical data to back this up but the source was reliable.

System volume and flow rate are critical in heat pumps and the buffer deals with both these issues in a simple way that the LLH does not.

So much of the design of systems is stemming from large sales organisations that need to simplify matters for their sales teams by coming up with a 1 size fits all package. 1 of these organisations was with us yesterday and freely admitted that they do not optimise performance, being driven rather by margin and competition. As Big Freeze has said before look to what the northern Europeans do with heat pumps for the best steer.

Jon

TiredGeek
29-04-2011, 08:35 AM
Thank you guys, it's good to get feedback and advice :)

Bigfreeze: taking the flow setters out is easy, but I am not a professional and wouldn't know where to start to estimate flow rate by the flow temps :)

MikeHolm: the flow setters are the ones Mitsi supply with the 14kw EcoDans. I did ensure they have more than 6" of straight pipe fore and aft, I figured turbulence would throw the readings off...

Jon: many good points, thanks.
I don't think it's a system volume size for me though, we have 30 rads on the system, most of them doubles, and 95% of the pipework in the house is 22mm with only the odd small run in 15mm. ALL the pipes around the ASHP and for the first 1/4 of the rad run is 28mm. In all, we estimate 230litres in there, surely enough.
While TRVs are fitted to most rads, they are all set at 5 so non should be closing, and even when we manually turned off 50% of the rads we still got 25L/min flow back from them.
The LLH was fitted to hopefully improve flow through the Mitsis but it seems like the FS and flexis are restricting the flow too much.
I've had the same advice re defrost stopping when demand for heat is lost, and probably from the same reliable source, but this is categorically not what has been happening to mine as I have witnessed failed defrosts and the stat was definitely still asking for heat.
Step controller: one was originally specced and paid for but then the company decided it wasn't required and simply installed them in parallel, refunded the money, eventually.
I suppose in the end a buffer tank (or 2 as I have two separate circuits for the rads and ASHP now, so 4 hot and 4 cold pipes) may have to be fitted, but I don't want to have to heat a further 120litres (or 240) of water every time the heating comes on, and with as much water in the system as I have I don't see it's actually going to do any good.

I think all the problems are flow related through the ASHP.

Either that or I have been miss-sold a system not capable of meeting the criteria we laid down at purchase time, simple enough: be able to do all an oil system can do. ie: economically heat the house to 20'c when we're there, and work on frost 'stat at 10'c when we're not. So far it's just about doing the first, but not cheaply, but massively failing to do the second!

Bigfreeze
29-04-2011, 04:23 PM
TiredGeek,

Take out the flow setters, they're not needed. You can check your inlet outlet temps to determine if you have enough flow. You should be between 4-7C difference if all is ok. Personally I think the pump is undersized too.

I mentioned to you before that the buffer was the better option and I think you should consider it, especially seen as its rads you're heating.

Imo you were sold a pup, two systems that should never be seen together especially in an old poorly insulated house are heatpumps and rads. The fact they're air to water only makes things worse. If I were you I'd do the following, plan on installing a buffer with an oil boiler as a 2nd level back up when the temps go below freezing. Keep one AS unit and tell the installer to remove the second and that if he doesn't reimburse you for it you will take him to court and clean him out. As you can see from another thread in this section that there is already precedent for this.

TiredGeek
30-04-2011, 06:58 AM
Bigfreeze, in the proccess of doing pretty much that :)
We've asked for a refund, had advice from Trading Standards, and are putting a case together of all the emails and brochures and other stuff.
That Don't get done get Dom episode helps as well as we were promised low running costs and we still have our original sales blurb to quote back the "up to 75% cheaper to run" bit :)

We put the LLH in because we had it and if it had worked it wouldn't have mattered too much if we failed to get money back.
Buffer tank, for my plumbing set up I'd have to put in two of them, not enough connections on one, so I'd end up with 240 extra litres of water to heat!

Jon Glanfield
26-05-2011, 05:23 PM
Hi Tired Geek I meant to post a while ago but have been busy. Once it was clear from your posts who the design company was, I gave them a ring to chivvy them up on your behalf, in dealing with your complaint. I sincerely hope by now they have been in touch, but if not PM me and I'll confirm the details of the discussion we had and who was to be contacting you.

Kind regards

Jon

yinmorrison
26-05-2011, 11:51 PM
A low loss header should have a maximum velocity through it of 0.5m/s so a 42 mm ' header' is not a header at all.It should be between 300m dia and 350mm dia by about 1 to 1.5M long. In other words it is acting like a Buffer vessel or a storage vessel , whatever you want to call it.You have 2 x 14Kw units so each require pumps selected at the max flow not the minimum, and should be 25-55 pumps on each primary circuit which have a max head available of about 65Kpa at that flow rate. I notice that Mitsi do not provide head loss through their unit which is strange to say the least as they must be plate heat exchangers on the Ecodan. In any case, as stated previously a Buffer/Storage Cylinder should be highly recommended on this type of system.

TiredGeek
30-05-2011, 08:05 PM
You have 2 x 14Kw units so each require pumps selected at the max flow not the minimum, and should be 25-55 pumps on each primary circuit which have a max head available of about 65Kpa at that flow rate. I notice that Mitsi do not provide head loss through their unit which is strange to say the least as they must be plate heat exchangers on the Ecodan. In any case, as stated previously a Buffer/Storage Cylinder should be highly recommended on this type of system.
We used the pumps specified by the company who supplied the units, but now we are of the opinion they haven't a clue ;)
Ongoing saga, wish we'd never gone this route. 20/20 hindsight is a real bummer :)