I can see that works when there is no HEX.  Basically if the rads are wide open the low impedance route is through the rads, if they close then the buffer acts as a bypass valve (many advocate putting a pressure operated bypass valve at the far end of the rad circuit, this is an alternative).

But when there is a HEX which is likely to have a constant high impedance, I cant see it, unless you put it on the rad side of the HEX not the HP side (which is where yours currently is) in which case it may well work as Kendra intend.

Personally I still wouldn't do it as it depends on a pressure balancing act.  As Kendra say its likely better than a 4 pipe arrangement, but how is it better than a simple volumiser in the return?  I willing to be convinced though if someone can come up with an argument.

Posted by: @iantelescope

↑

@jamespa 

 

Thanks James.

Positioning  a two port Buffer across the Radiator circuit is promoted by Kendra in his excellent  "Buff up your knowledge of buffer tanks "

see attached pdf.

This presumably only works where the parallel pipes have a lower resistance compared to the Buffer itself.

Kendra_pdf.pdf

If, when using a Heat Exchanger , the Buffer resistance was increased , the Water would flow through the Heat Exchanger and the Operational volume would be increased by 50 l.

 

 

ian

In the Kendra example, note the position of the circulation pump, which is after the buffer tank, and I suspect would only work with any degree of efficiency if there is also a water pump within the heat pump.

Assuming that any zone valves or TRV's are after the circulation pump shown, then if these valves close then the water can no longer flow through the radiators or the buffer tank. The only way that water could flow through the buffer tank is if there is also a water pump within the heat pump or the flow or return pipe work prior to the buffer tank.

For the Kendra system to operate at maximum efficiency the flow rate through the heat pump water pump would have to match the flow rate through the circulation pump, otherwise there will be some water flow and hence mixing taking place within the buffer tank.

@jamespa we get over 15% efficiency gain without a buffer.

I've commented on this thread a number of time, if you continue to believe the dinosaur hydronic pushed by manufacturers and certain people, then carry on.

If you don't want to listen and even acknowledge what I have to say, who has been installing heat pumps for over 15 years.

Then Im out of here.

Regards 

Ken

• Posted by: @ken-bone

  ↑

@jamespa we get over 15% efficiency gain without a buffer.

I've commented on this thread a number of time, if you continue to believe the dinosaur hydronic pushed by manufacturers and certain people, then carry on.

If you don't want to listen and even acknowledge what I have to say, who has been installing heat pumps for over 15 years.

Then Im out of here.

Regards 

Ken

Sorry, but I don't believe in buffers (unless there is a proven need) and there won't be one in my system.  OP asked where to put a buffer in his currently mis plumbed system.  I suggested he turns it into a volumiser not a buffer because then it does no material harm afaik.  Personally I'd leave open or take heads off some trvs and omit the buffer, which is exactly what I suggested earlier in this thread.

@jamespa 

Capacitors , Water and electrical:

The Purpose of the Tank is to smooth out the variations in the Water Temperature by increasing the Tank volume.

The Tank should act as an Integrator. An integrator should be fitted in parallel.

A comparison with Electrical Capacitors:

In Electronics , an Electrical Tank , or Capacitor, is used to Smooth out, or Integrate,  the variations in voltage at the output of a power supply.

In Electronics . Smoothing Capacitors are always fitted across the source to smooth out the variations in DC voltage.

Capacitors in series are used as Differentiators , blocking DC but allowing AC variations through.

So, for Electronics engineers to fit Capacitors , or tanks in series is heretical!

Water Flow Electrical Current flow

The Power flowing through the Heat Exchanger is :

       Power _through_Heat exchanger_primary  = SH X Temp difference across Tank X Water_Flow _rate_through_Heat_Exchanger

Now, the Water_flow_rate through the HEX_primary depends , as you say, on the relative resistance of the path through the HEX

and the resistance of the Tank.

I can therefore Increase the Tank Resistance by adjusting the switch to act as a bleeder , increasing the Water Flow through the HEX.

A parallel arrangement is far easier to physically implement than an in series arrangement.

Freezing tank

Fitting the freezing tank in Series with the HP return is even more difficult.

The visiting French Samsung Engineer admitted that a "in series tank " would be very difficult to implement, finally accepting the current parallel arrangement.

ian

Posted by: @ken-bone

↑

@jamespa  which if you read my comments, I also suggested this.

Good, we agree then!

The problem is to persuade @iantelescope and, more to the point, his installer.  My personal theory (based on what @iantelescope tells us of the history) is that Samsung suggested adding a volumiser because of short cycling, but the short cycling was due to poor pump adjustment and/or too many rads shutting down.  So a volumiser/ buffer was added wrongly, making the short cycling worse.  The correct solution would probably have been simply to open up some rad valves and amend the WC curve.

@iantelescope

I afraid that your analogy of a buffer tank being the same as a capacitor is flawed. For a buffer tank to be similar to a capacitor it would have to have a membrane separating the top half from the bottom half. I can explain in greater detail if you wish.

@iantelescope

I have worked with French engineers, I was not impressed.

@derek-m 

Sorry, Derek, I was cut OFF following a flurry of Emails .....

I hope I have not started a personal dispute ...............

Capacitor ,Tank?

I know that a Capacitor has a membrane , and that a Tank does not.

My point is that both a Capacitor and a Tank are Integrators.

As Integrators, both a Capacitor , and a Tank act as low pass  Filters.

The Tank , acting as a Filter, removes the High frequency components , smoothing, and extending the Water Cycle and extending the cycle time.

The extension of cycle time is shown in the MCS calculation of the Tank Volume.

ian

Im not sure how helpful the analogy is.  The key point is that the buffer is (at DC) most probably a relatively low impedance path compared to the heat exchanger, and thus the current passes primarily through the buffer.  Thats not what you want.

@iantelescope

It may be beneficial to understand why a heat pump may cycle, and then look at ways in which the cycling may be stopped or at least reduced.

Heat pumps cycle (switch on and off) when the amount of thermal energy that they are supplying exceeds the demand, which causes the LWT to increase above that required, until it reaches the upper limit and stops the compressor. So to stop or reduce the cycling it is necessary to reduce the thermal energy output and/or increase the demand.

The standard way that a heat pump reduces its output is by slowing down the compressor, which is normally achieved by setting the desired LWT to a lower value. This works until the compressor reaches its minimum operating speed and cannot go any lower, so if the demand is still lower than the supply, the LWT will eventually increase until it stops the compressor. The compressor will restart when the measured LWT drops to the minimum limit, which is often 5C below the desired LWT. How long the compressor is stopped is therefore determined by how long it takes for the LWT to cool, or may also be set by a time delay within the controller, which prevents the compressor from being restarted even though the LWT has fallen to the lower limit.

So what can be done to stop or at least reduce the frequency of compressor starts?

One way would be to increase the thermal energy demand by requesting a higher indoor temperature.

It may be possible to increase the length of any internal delay within the controller. This would have a dual effect of not only delaying the restart of the compressor, but also allowing the LWT to cool to a lower temperature before it is reheated, thereby increasing the running time.

A further method is to increase the water volume of the system, such that even when supply exceeds demand, it will still take longer for the water to be heated and then cool once. 

Because your system has the PHE, the total system volume is both the primary and secondary, but that is only true if the secondary water pump remains operational when the compressor has stopped, otherwise the only active volume is that of the primary circuit, and then only if the primary pump remains operational or at least operates periodically to distribute the heat around the system.

It would be useful if you could provide details of the settings on your system, what switches the two pumps on and off, if you have any TRV's on your system and if they could be limit the flow through the radiators. Are you operating your heat pump in WC mode and if so what are the settings?

Finally monitor the operation of your system noting when the compressor and pumps stop and start and the various temperatures at which this occurs along with the timing of starts and stops.