Paddle or Shoe Valves? (Now Ball Valves)

[le-cerveau]

Whilst I am still working out some of the more difficult questions:

 

Does anyone know if the claimed no leaking ‘paddle’ type valve is better than a shoe valve?

Pictures taken from Danfoss literature:

I am looking to utilise 3-port valves to control my heating/cooling system so it has to isolate the UFH from the TS when there is no heat call (constant circulation) and also isolate the TS from the ASHP when cooling.
All literature on valves shows some leakage but the Danfoss literature claims the ‘Paddle’ valve seals 100%.

As anybody got experience in these and does anybody else make ‘paddle’ type valves?

Edited June 14, 2016 by le-cerveau
Pictures re-added, Title updated for Ball valves.

[ProDave]

What's your reason for 3 port valves?  Yes I agree they are quite clever the way they work, they also give a lot of trouble and in a multiple valve system designing the controls for them can be interesting.

I prefer individual 2 port valves for each individual circuit, a lot more reliable and less complicated. You can also get normally open (energise to close) 2 port valves if that helps.

[le-cerveau]

My design has 3 x 3-port valves, I could replace with 2-port valves but that would mean 6 valves so I was reducing the number of moving parts!

The controls will be bespoke anyway as nobody does an off the shelf system that does heating/cooling and Thermal Stores (that I know off), my design is based loosely on Jeremy's, rough layout (at the moment) below. 

Edited June 10, 2016 by le-cerveau
changed 2 valves to 3

[PeterW]

OK so if I have this right :

ASHP and Boiler can heat the TS directly

TS feeds the UFH directly for heat

ASHP can cool the MVHR via the duct cooler

ASHP can cool the UFH by the Heat Exchanger..???

You would need to properly isolate the TS from cooling as I would expect the pressure differential of pushing cold through a PHE and a fan coil is much higher than the TS and the 3 way valve.

Also any reason for only letting the gas hit the top section of the tank...?

 

 

[le-cerveau]

Peter

7 hours ago, PeterW said:

OK so if I have this right :

ASHP and Boiler can heat the TS directly  YES

TS feeds the UFH directly for heat YES when 3-port valve '1' is open

ASHP can cool the MVHR via the duct cooler YES

ASHP can cool the UFH by the Heat Exchanger..??? YES at same time as MVHR, 3-port Valve '2' & '3' change to isolate TS from Circuit and ASHP respectively.

You would need to properly isolate the TS from cooling as I would expect the pressure differential of pushing cold through a PHE and a fan coil is much higher than the TS and the 3 way valve.  Hopefully the 2 x 3-port valves do that?

Also any reason for only letting the gas hit the top section of the tank...?  Usual way of charging a stratified TS

 

 

Tried to answer directly to questions.

I am trying to keep the number of motorised valves to a minimum by using 3-port and 1-way valves, hence the question about paddle/shoe valves.

An alternative would be to replace the 3 x 3-port valves with 6 x 2-port valves.

[PeterW]

OK - so I can't see the point in the heat exchanger in the circuit as you are better off just letting the cold circulate to the UFH as it won't really have that much capacity in the system and there seems to be no circ pump on the heating side of the H-Ex and I don't think the UFH will pull enough to keep a decent flow.

Why two UFH manifolds..? Can you not get all the circuits off one..? I'd also consider adding the MVHR coil as a circuit to the UFH manifold.

Losing the H-Ex I think I can get it down to 2 x 3 way valves which would be on the ASHP - Hot and Cold - and 2 x 2 way on the feed from the TS to the UFH.

In heating mode, ASHP would be to A, feeding into the TS and the UFH >> TS valves would be open.

In cooling mode, ASHP would be to B, UFH >> TS valves closed and chilled water would be circulated to the floor and the MVHR coil.

Would that work for you..?

 

[Nickfromwales]

With an ashp there's usually antifreeze inhibitor in the primary water so a PHE can be used to separate the internal and external bodies of water ( to reduce the volume of treated water that's needs to be refreshed when the antifreeze dies off.   )

 

[le-cerveau]

Peter/Nick,

it is not a PHE but a hydraulic separator, to separate the ASHP pump for the UFH manifold pumps, ASHP manufacturers instating (they have looked at the layouts).  That should overcome the cool pull problem, the Manifold pumps also have the job of pulling from the TS.

There are two manifolds as there is 4km, 20-24 ports worth, of UFH piping (400m2 house).

Have considered a MVHR preheat coil, would just put a pull from the TS with its own pump and thermostat and let it get on with it by it's self.

The final valve layout will most likely change before it is finalised, I suppose the ultimate question is: 3-port valves (paddle) that should provide a 100% seal or multiple 2-port valves.  I know Jeremy went with 2-port for the 100% seal vs 3-port leaking?

 

[Nickfromwales]

I'm a 2-port man myself, ( unless it's an old arrangement with boiler / pump overrun   )

 

[PeterW]

7 hours ago, le-cerveau said:

Peter/Nick,

it is not a PHE but a hydraulic separator, to separate the ASHP pump for the UFH manifold pumps, ASHP manufacturers instating (they have looked at the layouts).  That should overcome the cool pull problem, the Manifold pumps also have the job of pulling from the TS.

There are two manifolds as there is 4km, 20-24 ports worth, of UFH piping (400m2 house).

Have considered a MVHR preheat coil, would just put a pull from the TS with its own pump and thermostat and let it get on with it by it's self.

The final valve layout will most likely change before it is finalised, I suppose the ultimate question is: 3-port valves (paddle) that should provide a 100% seal or multiple 2-port valves.  I know Jeremy went with 2-port for the 100% seal vs 3-port leaking?

 

Ok so two manifolds is probably the right answer ..!! 

You can get low powered valves with 12-24v actuators that may do what you want - look for ones suitable for air usage too as they normally seal 100% by design. 

Downside is usually there is a smaller orifice so unlike a 22mm valve having an 18mm orifice, these only have 8-10mm 

Plenty of examples here 

http://www.solenoid-valve.world/by-material/brass-solenoid-valve/hvac-actuated-valve?gclid=CjwKEAjwkPS6BRD2ioKR7K245jASJAD1ZqHOK0Rj6MMhSry7z0TP1EEyiA3d6Nd523N2uEo97YLJphoCNKXw_wcB

[Jeremy Harris]

I used 2 port motorised ball valves.  100% seal, unlike the paddle or shoe valves; I tested a paddle valve and it did let by a tiny fraction under pressure, and as I wanted 100% isolation between the warm and cool sides of our ASHP pipe work was not prepared to risk using a paddle valve.  I also bought a three port motorised ball valve to test, but that was very restrictive, as it uses a slot in the ball as the feed to the ports, rather than a hole right through.

I bought a box of DN20 12V motorised valves, having bought one sample, taken it apart and found it to be well designed and made (it came from China).  I also bought some similar DN15 valves.  The DN20 valves are just like a normal DN20 ball valve body, with 3/4" BSPF threads in the ends, but they have a small geared motor, a relay and two micro switches as limit switches inside the housing on the top.  The ones I have are three wire types, one wire to ground and apply power to each of the other wires to open or close the valve.  As soon as the valve has opened or closed it stops drawing any power, as the limit switch operates.  They only draw around 100mA max at 12V for the two or three seconds the motor is running.

The ones I have use standard small micro switches and a standard relay internally, but the motor and gear train seemed to be specific to the valve.  This is why I bought a box of them, as I have loads of spares.  In all probability the only thing that might fail is the motor unit, and it's ten minutes work to replace this.  I paid about £8 each for them and bought 10 altogether, although I'm only using 2.  Here's a photo of a spare 12V DN15 three port valve that I happen to have as a sample here, the ones that have the restrictive flow (the two port valves have a straight through bore that's the same as a standard ball valve when open):

 

[le-cerveau]

So, I think I am convinced it should be ball valves! (simple answer)

If a ball valve is available with decent 3-way flow would people recommend that that is a viable option, I am still keen to reduce the number of moving parts as far as possible, because with 2-port valves I double the number of motorised valves? 

[Jeremy Harris]

When I was looking around, I couldn't find any full-flow three way ball valves, and looking at how they have to be constructed internally in order to work, I doubt that there are any.

My preference would be to always use two way valves, but that's largely driven by a prejudice of three way valves borne out of having had several fail on different heating systems over the years (every house we've ever had with a three way valve has had a valve failure - maybe we've just been unlucky!).

Is there any way you can reduce the number of motorised valves needed?

I only have two, one on the return from the UFH manifold, to isolate it when there is no call for heating and one on the flow to the buffer tank/preheat to isolate that whenever there is a call for cooling. 

I have the absolute minimum number of valves in the whole system, and only fitted isolating valves where it was essential, so there are none at all anywhere in the pipework to the buffer tank, or even on the heat pump flow and return, as the volume that flows from the primary circuit drain is so low that I can catch it all in a large bucket.   The UFH pipes remain full of antifreeze when the heat pump is drained down and so the volume is just that in the pipes and in the buffer tank coil.

[le-cerveau]

I need to look at the current layout and see how I can minimise (if possible) the number of valves.  One problem is that the ASHP producers have driven me to include a hydraulic separator in the circuit when cooling, to separate the ASHP circulating pump and the 2 x UFH manifold pumps, also I have a duct cooler on the cooling side.  I will have a think about it but nothing is set in stone yet, net event the TS type (or option for Buffer tank and TS).

I think the key thing is I will be insisting on ball valves, and I will keep researching the 3-port option (with the manufacturers).

[PeterW]

1 hour ago, le-cerveau said:

...... One problem is that the ASHP producers have driven me to include a hydraulic separator in the circuit when cooling, to separate the ASHP circulating pump and the 2 x UFH manifold pumps.....

What have they specified exactly as I'm slightly lost on their logic ! Do they mean a low loss header..?? 

[le-cerveau]

Peter, yes a hydraulic separator/Low Loss Header

[le-cerveau]

Update on Ball valve flow:  Straight through the 'T' if is at full capacity but round the 'L' it is approximately 27%!!!!! so I guess it will be 2-port valves for me!

Plumbing design update required!

[PeterW]

I'd look at flow rates before you discount them - are these 15 or 22 mm valves ..?

Flow rates from ASHP are quite low as are UFH which is why I queried the low loss header use 

[le-cerveau]

They are available in ½", ¾" and 1" with a 12mm, 15mm and 20mm orifice respectively with a Kv (m3/h) of 8.6, 21 and 26.

The valves I am currently looking at are by JP Fluid Control http://tameson.co.uk/electric-ball-valve/

So a 1" valve with a 20mm orifice has a Kv or 26 m3/h (straight) and 7.02 m3/h ('L' diversion)

Definition Kv-value: Amount of flow (m³/hour) of water of 20°C in a valve with a pressure loss of 1 bar.

The amount of flow of a fluid through a (solenoid) valve can easily be calculated with flow coefficient Kv. The Kv-value expresses the flow rate of water in m³/hour in a valve with a pressure loss of 1 bar and a temperature of 20°C.

 

where:

• Q = flow rate of liquid (m³/hour)

• Kv = flow coefficient (m³/hour)

• SG = Specific Gravity (=1 for water)

• dp = pressure differential over the valve (bar)

The manifolds have a max flow of 30 l/min which is 1.8 m3/h, so in theory the 3/4"(15mm) one with 21 m3/h / 5.67 m3/h would suffice.

More food for thought.

 

[PeterW]

What's the flow rate of the ASHP...? That will be the limiting factor in this I expect

[le-cerveau]

Minimum flow rate 12 l/min, max 46 l/min which is 2.76 m3/h so the ¾" (15mm) one would suffice but if I went this way I would probably go for 1" (20mm) valves to reduce flow pressure.

[le-cerveau]

I have done a complete control re-design, reduced to 2 motorised valves by utilising the Hydraulic separator/Low Loss Header (LLH) as the bypass and then switched to 2-port valves with bypass valves to control the flow.

The way I think it will work (not being a fluid dynamics expert) is a follows:

House Calls for Heat: Valve 1 & 2 open.
Water circulates round the UFH and through the TS (via mixing valve) to provide heat to house, the bypass valve prevents water flowing to the LLH.  If the TS calls for heat it flows from the ASHP, again the bypass valve prevents heat going to the LLH.

House at Temp: Valve 1 closed, Valve 2 open.
Water continues to circulate, but Valve 1 is closed so the bypass valve open and water circulated through the LLH.  The pump on the MVHR circuit will prevent water passing that way and the one-way valves at the ASHP prevent it short-circuiting through the TS and via the mixing valve.  The TS can still call for heat as before.

 House calls for Cool: Valve 1 & 2 closed.  Water circulates through the LLH as above, the ASHP runs in cooling mode and provides cool water to the LLH via the bypass valve, MVHR duct cooler can also be energised to provide some air cooling.  TS call for heat is ignored due to NC relay opening on cool call.

I have just shown the control elements for simplicity and which way the pumps work.  (pdf version also available for clarity)Heating-DHW-Design(2-port)-Control.pdf

How does my logic stack up?

[PeterW]

Looks OK ...! Are you blending down the water to the UFH for a reason from the TS..? The manifolds will do this anyway, could be another component to lose..? Also, a pump won't stop flow - if you really want to isolate the duct unit it will have to be via a valve.

Looks pretty neat and simple though !

 

[le-cerveau]

Peter, Not overly worried about some seepage into the MVHR duct heater in heating mode as it will be a: minimal and b: of such a low temp as not to make a significant difference to the MVHR, but will consider isolation valve there also.  As to the blending, the UFH only needs very low temp maybe 24C, I was pre blending to reduce heat takeout from the TS, I was considering losing the mixers from the manifolds (2 of 1 upstairs, 1 downstairs) but thought it would give me another layer of control (complication) to run the downstairs manifold slightly warmer than the upstairs (as heat naturally rises), I probably should lose one or the other, to reduce components.

I still have time but the design is coming together.

[PeterW]

I think you may struggle to get a proper temperature / flow on the UFH if you lose the manifold mixers as the whole of the 3 UFH circuits will become a homogeneous flow that has little control.

You can feed TS water at 65c to the manifolds - they will only take what they need and then circulate the rest.

Slightly off the wall - excuse this one - but if you are only looking for minor heat/cool from the duct heater, can you not run it as a separate circuit on the UFH.?? Downside is you can't give it very cold water in the summer as the floor won't be below 16 or so, but is it worth a thought..?