Basic Heating Control Understanding Help

[Shaun McD]

Hi all, getting closer to signing myself up for the plumbing and heating on our new build but have some basic heating system operations that I would like to clarify/confirm to make sure I am not missing out on anything!

 

So basic system;

Mono bloc ASHP

Appropriate ASHP suitable DHW Cylinder

Assuming no buffer tank

UFH coming from 2 manifolds;

- both manifolds will have pumps and blending valves

 

Questions;

1. How does ASHP know the DWH needs heat? I assume it connects to a probe in the cylinder?
2. How does ASHP know the heating system needs heat? Is this via a temp probe somewhere in the house connected to ASHP?
3. If the above is correct, I assume the ASHP uses this and the outside temp to work out temp of water to output based on weather compensation?
4. Does each manifold need to have a motorized valve to disconnect it from heating circuit? (Mainly curious if this is best practise, but in reality I plan to opt for this to allow for some sort of solar gain distribution like Jeremys) 
5. It looks to be standard practise to have a stat and valve for each manifold loop, and if any of the loops is calling for heat then this is the trigger to run the pump on this manifold?
6. If I opted to run without stats in each room, how would the manifold pumps be triggered? (ignoring whatever solution is used for solar gain distribution) Would this be linked to the trigger in point 2 above?

 

Thanks in advance for input and no doubt follow up questions!

[ProDave]

1: Most ASHP's have a temperature probe that goes into a thermostat pocket on the tank.

 

2: Most ASHP'will have some form of call for heat / thermostat input.

 

3: Weather compensation will lower the temperature of water to the UFH when it is less cold outside to get the best COP from the ASHP.

 

4: I would personally have a 2 port valve for each manifold and one for the DHW tank.

 

5: you have a room thermostat for each zone, not each loop, a large room or zone will be multiple loops.  All dealt with by the manifold control box.  An alternative school of thought on a very well insulated house where the heat input is low, is don't have individual zones, have the whole floor as one zone.

 

6: I would always have at least 1 room stat per floor.

 

The problem with all these general answers, is the details of exactly how each ASHP communicates and deals with the other heating components is not a standard thing and it varies a lot from one make of ASHP and the next.  So expect the detailed design to be very ASHP dependant.

[SteamyTea]

8 minutes ago, Shaun McD said:

Assuming no buffer tank

May be a bad assumption, it depends on the water volume in your heating system is.

9 minutes ago, Shaun McD said:

If I opted to run without stats in each room, how would the manifold pumps be triggered

Depends on your ASHP, they can calculate energy required from flow, return and OAT temperature difference.

I assume you have already had a full, room by room, heat loading calculations done and you know tour pipe spacing for the lowest flow temperature you can satisfy the overall heat load.

[JohnMo]

No buffer equals big zones only, so possibly only one zone per floor, but you need to calculate the water volume for each zone to ensure it meets your specific heat pump requirements. I have a small zone (2x50m loops) the thermostat can switch the zone off but not on, as the volume is way to small to run on its own, without a buffer.

 

1. temp probe, when temp hits a defined limit, the ASHP drives a 3 port valve to DHW and ramps up flow temperature.

2. My ASHP gets a trigger from the single house thermostat, call for heat ASHP starts, no call ASHP goes into standby mode

3. Not sure your assumption is correct. The WC curve does not take anything from the thermostat or room temps. It's a basic, when external temp is X, flow temp will be Y, straight line curve.  Even if it did UFH is way to slow to react to take room temp in to account. It just chases its own tail (been there, done it)

5. If you want a stat on each loop you need a buffer sized for the smallest loop.

6. You could do away with the wiring centre and control the pump directly by the thermostat.

 

I dumped the manifold pump and mixer, as they incurred a temperature drop across the mixer of 3 to 4 degs, cooler but with two floors and different flow temps you maybe need one or both.

[Temp]

2. Generally with ufh and multiple stats there is a "wiring centre" at each manifold.

 

This has inputs for each stat and outputs to control the valves on the manifold.

 

It also provides a BE (Boiler Enable) which is the logic OR of all the stats. Eg if stat A or stat B or stat C... calls for heat then BE is active.

 

The BE signal can be used to control an ASHP or if there is a Thermal Store it controls the output of the Store.

 

On systems with just one or two stats

you don't really need a wiring centre but something else (wire and possibly a relay) performs a similar function.

 

The BE signal also controls the loop pump on the manifold and any two port valves between heat source and manifold. Sometimes the wiring centre provides separate outputs for these.

 

Edited May 28, 2023 by Temp

[JamesPa]

12 hours ago, Shaun McD said:

Hi all, getting closer to signing myself up for the plumbing and heating on our new build but have some basic heating system operations that I would like to clarify/confirm to make sure I am not missing out on anything!

 

So basic system;

Mono bloc ASHP

Appropriate ASHP suitable DHW Cylinder

Assuming no buffer tank

UFH coming from 2 manifolds;

- both manifolds will have pumps and blending valves

 

Questions;

1. How does ASHP know the DWH needs heat? I assume it connects to a probe in the cylinder?
2. How does ASHP know the heating system needs heat? Is this via a temp probe somewhere in the house connected to ASHP?
3. If the above is correct, I assume the ASHP uses this and the outside temp to work out temp of water to output based on weather compensation?
4. Does each manifold need to have a motorized valve to disconnect it from heating circuit? (Mainly curious if this is best practise, but in reality I plan to opt for this to allow for some sort of solar gain distribution like Jeremys) 
5. It looks to be standard practise to have a stat and valve for each manifold loop, and if any of the loops is calling for heat then this is the trigger to run the pump on this manifold?
6. If I opted to run without stats in each room, how would the manifold pumps be triggered? (ignoring whatever solution is used for solar gain distribution) Would this be linked to the trigger in point 2 above?

 

Thanks in advance for input and no doubt follow up questions!

I would add to the design requirements - get the system sizing right.  Not oversized (like many) and obviously not undersized.  If you don't efficiency or comfort will suffer.  Don't be tempted to 'go big' for safety, you will regret it. 

 

I presume that you have a good heat model and a good handle on ventilation rates?

[PeterW]

13 hours ago, Shaun McD said:

Assuming no buffer tank

incorrect assumption.. 🫣

 

put a 2 port buffer in, ASHP stat on the buffer and then carry on as previous.

[JamesPa]

13 minutes ago, PeterW said:

incorrect assumption.. 🫣

 

put a 2 port buffer in, ASHP stat on the buffer and then carry on as previous.

How do you reach that conclusion without knowing system volume and modulation capability of the HP?  

 

I agree it's incorrect to assume a buffer isn't needed, but surely it's equally incorrect to assume one is needed.  I do agree 2 port if you need one and pref in return not flow to avoid any reduction in flow temp.

[Nickfromwales]

27 minutes ago, JamesPa said:

How do you reach that conclusion without knowing system volume and modulation capability of the HP?  

 

I agree it's incorrect to assume a buffer isn't needed, but surely it's equally incorrect to assume one is needed.  I do agree 2 port if you need one and pref in return not flow to avoid any reduction in flow temp.

Flow temp will be affected either way, as the cooler return will reduce the output of the ASHP, so that's just moving the 'problem' from one place to another imo. Truth is it's not actually a 'problem', and if it's within the heated envelope all latent losses go towards the target temp anyhoo. 6 and two 3's.

[SteamyTea]

1 minute ago, Nickfromwales said:

all latent losses

Sensible losses.

Latent, in thermodynamics, which this is, is when the phase changes.

Sensible heat is when the materials heated stay in the same form.

[Nickfromwales]

Just now, SteamyTea said:

Sensible losses.

Latent, in thermodynamics, which this is, is when the phase changes.

Sensible heat is when the materials heated stay in the same form.

Yes, boss.

 

Just the words "me" and "sensible" don't seem to work in the editor here. Keeps auto-deleting.

[Shaun McD]

14 hours ago, ProDave said:

1: Most ASHP's have a temperature probe that goes into a thermostat pocket on the tank.

 

2: Most ASHP'will have some form of call for heat / thermostat input.

 

3: Weather compensation will lower the temperature of water to the UFH when it is less cold outside to get the best COP from the ASHP.

 

4: I would personally have a 2 port valve for each manifold and one for the DHW tank.

 

5: you have a room thermostat for each zone, not each loop, a large room or zone will be multiple loops.  All dealt with by the manifold control box.  An alternative school of thought on a very well insulated house where the heat input is low, is don't have individual zones, have the whole floor as one zone.

 

6: I would always have at least 1 room stat per floor.

 

The problem with all these general answers, is the details of exactly how each ASHP communicates and deals with the other heating components is not a standard thing and it varies a lot from one make of ASHP and the next.  So expect the detailed design to be very ASHP dependant.

 

Thanks @ProDave, why always a 2 port valve for each manifold and DHW? (ignoring solar gain redistribution). Would the one for DHW not be surplus to the valve controlling heat V DHW?

House is single storey but in 3 connected blocks so hoping I can treat it as a single input for heating, and allow for solar gain redistribution within each manifolded area, i.e. east side of house & west side of house.

Appreciate your point on general answers, learning from this thread and others that there is more difference in ASHP controls between manufactures than I was expecting!! I will be sure to get into more detail once we have downselected the manufacturer we want (can get)! 

 

14 hours ago, SteamyTea said:

May be a bad assumption, it depends on the water volume in your heating system is.

Depends on your ASHP, they can calculate energy required from flow, return and OAT temperature difference.

I assume you have already had a full, room by room, heat loading calculations done and you know tour pipe spacing for the lowest flow temperature you can satisfy the overall heat load.

 

Tbh I am struggling to make sense of any solid calculations to know for sure which is the ideal spec, but I have used Jeremys spreadsheet for total envelope loss, and spent some time in loopcad doing this by loop etc. (I do need to fine tune this)

Re. pipe spacing, while I appreciate it may not be strictly needed and not be the most cost effective route, I am struggling to see why I would not default to 150mm pipe spacing;

1. It will allow for larger volume of water so that either manifolded area would be less dependent on the need for a buffer
2. My understanding is the more water in the system available to output heat, the lower the water temperature can be
3. With the kore insulated raft foundation, the is much larger than standard volume of concrete, acting making me think that having a large heat emitter running at low temps is better? 

These may not be great reasons, but on the otherside, I dont know what the downside to taking a cautious approach to this is, outside of cost of materials, labour at installation, is there any drawback to smaller pipe spacings?

 

14 hours ago, JohnMo said:

No buffer equals big zones only, so possibly only one zone per floor, but you need to calculate the water volume for each zone to ensure it meets your specific heat pump requirements. I have a small zone (2x50m loops) the thermostat can switch the zone off but not on, as the volume is way to small to run on its own, without a buffer.

 

1. temp probe, when temp hits a defined limit, the ASHP drives a 3 port valve to DHW and ramps up flow temperature.

2. My ASHP gets a trigger from the single house thermostat, call for heat ASHP starts, no call ASHP goes into standby mode

3. Not sure your assumption is correct. The WC curve does not take anything from the thermostat or room temps. It's a basic, when external temp is X, flow temp will be Y, straight line curve.  Even if it did UFH is way to slow to react to take room temp in to account. It just chases its own tail (been there, done it)

5. If you want a stat on each loop you need a buffer sized for the smallest loop.

6. You could do away with the wiring centre and control the pump directly by the thermostat.

 

I dumped the manifold pump and mixer, as they incurred a temperature drop across the mixer of 3 to 4 degs, cooler but with two floors and different flow temps you maybe need one or both.

 

Thanks @JohnMo I think my two manifold areas can be 120m2+ each (house 265m2), so im hoping that by taking the volume of water in the smaller of the two manifold zones and using that as the basis to require/size a buffer.

Note taken on WC, makes sense!

Can I ask how the pump and mixer was causing a drop in temp? Is the point of the mixer not to set the temp for all loops on that manifold? Is it in the response time that you seen an improvement then? And (assuming you have an upstairs manifold and a downstairs manifold) did you remove both pump & mixer from both manifolds or just one? Sorry if this is a simple question!

 

2 hours ago, Temp said:

2. Generally with ufh and multiple stats there is a "wiring centre" at each manifold.

 

This has inputs for each stat and outputs to control the valves on the manifold.

 

It also provides a BE (Boiler Enable) which is the logic OR of all the stats. Eg if stat A or stat B or stat C... calls for heat then BE is active.

 

The BE signal can be used to control an ASHP or if there is a Thermal Store it controls the output of the Store.

 

On systems with just one or two stats

you don't really need a wiring centre but something else (wire and possibly a relay) performs a similar function.

 

The BE signal also controls the loop pump on the manifold and any two port valves between heat source and manifold. Sometimes the wiring centre provides separate outputs for these.

 

 

Thanks for this @Temp this was my thinking based on previous experience with a gas boiler system, but I had assumed (again prob incorrect thing to do) that this would not be an ideal setup for a low energy build? I would have thought I wouldnt want to give any single room the ability actually turn on the ASHP in order to prevent short cycling?

My thinking on what would be expected for this type of system;

1. North facing room drops below room stat set point, opens manifold loop for room.
2. Manifold pump is triggered, pumping open loops in this manifold combined with now open north facing room.
3. If north facing room takes enough heat out of open loops then this will drop the temp in these zones, where a centralised house stat lives.
4. Only when this centralized stat drops below its set point (meaning that the whole manifold zone is at or below this temp) then will the ASHP trigger?

I was thinking the alternative would be to remove the room based stats completely and have either 1 centralized, or 2 (1 within a central point in each manifold zone) call the ASHP, assuming that the heat distribution within the zones will work itself out? Seems to be others on here who start out with loads of room stats but end up removing them.

 

2 hours ago, JamesPa said:

I would add to the design requirements - get the system sizing right.  Not oversized (like many) and obviously not undersized.  If you don't efficiency or comfort will suffer.  Don't be tempted to 'go big' for safety, you will regret it. 

 

I presume that you have a good heat model and a good handle on ventilation rates?

 

I have been using Jeremys calcs for whole house and then loopcad as a starting point for each "room", but I am thinking I will do a more exact room by room sheet alongside loopcad in the coming days!

 

1 hour ago, PeterW said:

incorrect assumption.. 🫣

 

put a 2 port buffer in, ASHP stat on the buffer and then carry on as previous.

 

@PeterW Am I right in that you suggest this as it is really the more fool proof option? Meaning that while it may not actually add any efficiency to a well designed system, it will help an otherwise poorly designed system? 

Second question, since the stat on the buffer is reading water temp v room temp, im guessing that this is either accounted for in the ASHP controls or do ASHP controls typically have input options to state that the stat is reading water temp v room temp? Suppose what I am wondering is does this make it a little bit trickier to set room temp, as there wont be a direct match from the buffer temp to the temp we want in room? If I want to set a room at 21c, I cant tell the ASHP to look at the buffer stat and heat it until it 21c right? This will be a different number...

Third question if I may, I have been looking at UVC's and looks like we have 3 priorities to hit, 1 - Coil size suitable for ASHP, 2 - Element location for PV (bottom of tank), 3 - Volume for our usage. I seen the, for lack of better phrase, "combi" cylinders which have both UVC and buffer in one, e.g. https://www.jouleuk.co.uk/pre-plumbed-heat-pump-package-w-buffer/ While this option looks to be a no go based on priority 2 (location of element for PV), assuming that such a tank exists that meets the 3 priorities and has an inbuilt buffer, are these a good option, or do you typically always opt for a stand alone buffer? 

Last question on this ha, any logic in having a buffer with an element connected to a secondary PV output?

[Nickfromwales]

2 minutes ago, Shaun McD said:

House is single storey but in 3 connected blocks so hoping I can treat it as a single input for heating, and allow for solar gain redistribution within each manifolded area, i.e. east side of house & west side of house.

I think it was Jeremy who mentioned just running the UFH pump to circulate the water in the floor(s) without the HP engaged. Great idea, and one I've adopted for my projects. I am using a dusk - dawn photocell to override the manifold pump(s) so they only come on when the sun is 'up', plus it runs for free then if PV is included.

[SteamyTea]

3 minutes ago, Shaun McD said:

I am struggling to see why I would not default to 150mm pipe spacing

Imaging two rooms in your house, both 3m by 4m, so 12m^2.

ONE is at the North East corner, with the 4m side running along the North wall.

The other is on the diagonally opposite corner, with the 4m running down the West wall.

The thermal losses, and gains, could be quite different.

 

[PeterW]

2 hours ago, JamesPa said:

How do you reach that conclusion without knowing system volume and modulation capability of the HP? 

System volume is reasonably irrelevant, you have to work on the minimum volume of the shortest loop, which even with a 100m 16mm PeX is around 14 litres. A heat pump that can modulate down to 1kW will still reach delta 7°C reheat in 8 mins so you’ve got a short cycle issue. Shorter loop, smaller volume and bigger issue

[SteamyTea]

3 minutes ago, PeterW said:

A heat pump that can modulate down to 1kW will still reach delta 7°C reheat in 8 mins so you’ve got a short cycle issue.

And that is assuming you want the full 1 kW.

My house takes about 400W to stay 15⁰C above OAT. 

[JohnMo]

2 hours ago, Shaun McD said:

Can I ask how the pump and mixer was causing a drop in temp? Is the point of the mixer not to set the temp for all loops on that manifold? Is it in the response time that you seen an improvement then? And (assuming you have an upstairs manifold and a downstairs manifold) did you remove both pump & mixer from both manifolds or just one? Sorry if this is a simple question

There is always a bleed through from the return to the supply, tried two different types of mixer, a Reliance and Ivar, both have the same outcome. Was wanting to flow 26 degrees into the UFH loops, but had to operate the weather compensation curve higher to achieve it, about 3 to 4 to degrees hotter. Which is rubbish.  If you are on the same floor, no need for mixers at all, in my opinion, others will not agree. 

 

Not sure where any benefit comes from a buffer with thermostat, as it cancels out running weather compensation. Just let it float on return temperature.

 

Loop spacing, extremely small difference in flow temp if your W/m2 is below about 20. At 20W/m2 your mean flow temp at 100mm spacing is 26 deg, while at 300mm centers it has jumped up to 28 deg. The further below 20 you go the smaller the difference.

 

 

[DanDee]

6 hours ago, Nickfromwales said:

Flow temp will be affected either way, as the cooler return will reduce the output of the ASHP, so that's just moving the 'problem' from one place to another imo. Truth is it's not actually a 'problem', and if it's within the heated envelope all latent losses go towards the target temp anyhoo. 6 and two 3's.

Whilst a cold system/house will take longer to reheat, so flow temp will slowing raise. The output of the HP is not affected, actually it goes full throttle putting out it's max capacity, and as the return increases, the HP will modulate down. 

 

The decision to have or not have a buffer is calculated on the type of emitters, system water volume, interior thermal mass.

 

An UFH screed or slab doesn't need a buffer and can take a wide range of the HP output.

 

 

[Shaun McD]

5 hours ago, Nickfromwales said:

I think it was Jeremy who mentioned just running the UFH pump to circulate the water in the floor(s) without the HP engaged. Great idea, and one I've adopted for my projects. I am using a dusk - dawn photocell to override the manifold pump(s) so they only come on when the sun is 'up', plus it runs for free then if PV is included.

 

Great to hear! In your photocell example, does that mean that when the ASHP runs at night you are not running the manifold pumps? 

 

5 hours ago, SteamyTea said:

Imaging two rooms in your house, both 3m by 4m, so 12m^2.

ONE is at the North East corner, with the 4m side running along the North wall.

The other is on the diagonally opposite corner, with the 4m running down the West wall.

The thermal losses, and gains, could be quite different.

 

 

Help me understand this, if each room has one large window then for sure solar gain will be different, but assuming the same fabric, losses will be pretty close to each other no? So since from a heat loss perspective would the emitter size not be the same? I appreciate if you can accurately map solar gain then you could decrease the emitter size appropriately, but not seeing the downside to keeping them the same?

 

5 hours ago, PeterW said:

System volume is reasonably irrelevant, you have to work on the minimum volume of the shortest loop, which even with a 100m 16mm PeX is around 14 litres. A heat pump that can modulate down to 1kW will still reach delta 7°C reheat in 8 mins so you’ve got a short cycle issue. Shorter loop, smaller volume and bigger issue

 

This is backing the concern with using room based stats at all, but I get stuck with the idea of maybe sometime for some reason liking or needing to have the ability to set room temps independently from each other, e.g. trying to keep a bedroom at 18/19c v centralized living at 21c. Think I will just have to come down on one side at some stage 

 

2 hours ago, JohnMo said:

There is always a bleed through from the return to the supply, tried two different types of mixer, a Reliance and Ivar, both have the same outcome. Was wanting to flow 26 degrees into the UFH loops, but had to operate the weather compensation curve higher to achieve it, about 3 to 4 to degrees hotter. Which is rubbish.  If you are on the same floor, no need for mixers at all, in my opinion, others will not agree. 

 

Not sure where any benefit comes from a buffer with thermostat, as it cancels out running weather compensation. Just let it float on return temperature.

 

Loop spacing, extremely small difference in flow temp if your W/m2 is below about 20. At 20W/m2 your mean flow temp at 100mm spacing is 26 deg, while at 300mm centers it has jumped up to 28 deg. The further below 20 you go the smaller the difference.

 

 

 

Interesting, yeah was pretty set on mixer valves based on reading here!! I guess I could still achieve the solar gain redistribution with a 2 way valve and pump. Would never have thought a bleed would have such an impact, and once you removed mixer you found 26c ashp output and 26c input at manifold? 

[JohnMo]

41 minutes ago, Shaun McD said:

once you removed mixer you found 26c ashp output and 26c input at manifold

Yes.

[Temp]

6 hours ago, Shaun McD said:

Thanks for this @Temp this was my thinking based on previous experience with a gas boiler system, but I had assumed (again prob incorrect thing to do) that this would not be an ideal setup for a low energy build? I would have thought I wouldnt want to give any single room the ability actually turn on the ASHP in order to prevent short cycling?

 

Yes that's why others have suggested one or two large zones rather than every room being one zone.

 

We have an oil boiler and they can't modulate down like gas and we wanted lots of small zone - so we fitted a Thermal Store as a buffer. That's not suitable for an ASHP though. 

[Nickfromwales]

54 minutes ago, Shaun McD said:

Great to hear! In your photocell example, does that mean that when the ASHP runs at night you are not running the manifold pumps?

Nope. The post refers to shoulder months where the heating may not be needed during the day, but instead where solar gain becomes problematic for E > S sunshine in the morning, with the S > W and N rooms then waiting to get the heat and balance out the slab / room temps.

 

If the house calls for heat the HP will be fired by the heating controls, as will the manifold pump(s). With my preferred technique of running the manifold(s) pump(s) during sunlight hours the problem of each far side of the dwelling behaving differently, heat wise, is reduced significantly.

[JamesPa]

6 hours ago, PeterW said:

System volume is reasonably irrelevant, you have to work on the minimum volume of the shortest loop, which even with a 100m 16mm PeX is around 14 litres. A heat pump that can modulate down to 1kW will still reach delta 7°C reheat in 8 mins so you’ve got a short cycle issue. Shorter loop, smaller volume and bigger issue

Sorry I was imprecise, I should have said/meant minimum system volume, ie the minimum volume that is guaranteed to be in circuit.  I guess that in many/most setups is the same as the minimum volume of the shortest loop

 

I confess that I had not realised how little water there is in a UFH system, if it really is that small then I agree a buffer (well volumizer, its not really a buffer in the sense the word is often used) is going to be necessary, if only to power defrost.

Edited May 28, 2023 by JamesPa

[JohnMo]

14 minutes ago, JamesPa said:

confess that I had not realised how little water there is in a UFH system

Not really a little water, there will be multiple loops, even with my 300mm centres I still have 28 litres just in the lounge, and there are another 7 loops..

[Nickfromwales]

6 minutes ago, JamesPa said:

if only to power defrost

This is a huge bone of contention.

 

If an ASHP has been spec'd to go into a dwelling that sympathises with such a unit, then freezing should be deleted? If it's Baltic outside and humid, ergo defrosting is prevalent, then the owner should be heating the DHW by the immersion(s) for that period of the year. The maths defo work out, as I've stated recently, just with the huge labouring of the HP being eradicated, plus the extension to the longevity of the HP / ancillary equipment. 

If you have solar PV then you will be banking electricity tokens all summer to 'spend' on this pursuit, but at the time where this will be most problematic your CoP will be 2:1 if you're lucky, and much much worse if defrosting is repeatedly required and added to the maths. 

In the arse of winter I recommend that you heat DHW via a cheap rate electricity period via the cylinder immersion and only do the daytime heavy lifting at most eg maintaining the absolute lowest set point of say 50^oC.

 

I have been oversizing UVC's for years now, for exactly this reason. Bulk heavy lifting when the CoP has the advantage, and boosting up the temps (heat energy capacity) via the immersion to stay away from having to do this whilst low temp heat is being produced. The oversize also helps massively during summer where extra DHW can be stuffed into the same device (by "overheating" the UVC with 'free' energy).