Advice on space heating passive house

[joe90]

11 minutes ago, Russdl said:

I really should have paid more attention at school

Every day is a school day (even fir old farts like me ?)

https://spark.iop.org/energy-series-and-parallel-connections#gref

[SteamyTea]

15 minutes ago, Russdl said:

At the risk of making myself look thick(er) can you explain how 3kW becomes 1.5kW?

Ohm's Law

 

https://www.calculator.net/ohms-law-calculator.html

 

 

[Russdl]

Thanks @joe90 ??
 

(bloody well hope I understand it!)

[Russdl]

@SteamyTea, thanks. 
 

(I hope no one is setting homework, it’s pub night tonight ?)

[Ajn]

20 minutes ago, Russdl said:

Not at all ‘obvious’ to me.
 

At the risk of making myself look thick(er) can you explain how 3kW becomes 1.5kW?

Each element has a fixed resistance, say 19.2 ohms for a 3kw element on 240v

using one on 240v gives a current of 12.5 amps (I = V/R)

using two in series the resistance is added so we have 38.4 ohms

So if used on 240v we have 240/38.4 = 6.25 amps

and the power W = V X A   gives 240 x 6.25 = 1500  or 1.5 kw.

[Radian]

It seems that people can be misled by thinking that if a 3kW kettle is OK using a three-pin 13A plug then any 3kW load is similarly safe. The kettle, however, is only operating for seconds and any warming of the lead and plug goes unnoticed as the heat has insufficient time to build up. But even a lesser load like a 2kW convector will cause noticeable warming of its cable and plug and this is when drawing only around 8 Amps. This is why typical 3kW immersion heaters are connected back to their own 16A mcb in the consumer unit.

 

I would be far more comfortable with 1500W running for twice the time - so long as it made up for losses and fit within any time-limited tariff.

[Ajn]

5 minutes ago, Radian said:

The kettle, however, is only operating for seconds and any warming of the lead and plug goes unnoticed as the heat has insufficient time to build up.

and those trades that borrow your extension lead but do not unroll it all and it comes back as one working but stuck together blob.

[joe90]

4 minutes ago, Ajn said:

and those trades that borrow your extension lead but do not unroll it all and it comes back as one working but stuck together blob.

Been there , done that. 

[joe90]

My immersions are run from individual fused spurs radial from the fuse board.

[ProDave]

7 minutes ago, joe90 said:

My immersions are run from individual fused spurs radial from the fuse board.

If you mean you have a Fused Connection Unit at the immersion heater, best replace it for a 20A double pole switch.

 

13A fuses get hot at close to 13A and in some makes of FCU they get hot to the point of destroying it when on for a long time.  If ever you smell a fishy smell check your FCU's for overheating.

 

If you have a 16A MCB at the consumer unit, no additional fusing is needed.

[TerryE]

Picking up @Radian's point you do need to be careful wiring devices that draw high currents for extended periods.  Contacts can start to heat up, this leads to oxidation which increases resistance which increases local resistance and this can escalate to melt-down.  This can be a particular problem with the flexible multi-core connector cables which should really be terminated with a decent crimped connector.

 

However the Willis is designed to take 12A for extended periods.  If you wanted to put 3kW into the slab using this approach the you'd need 2 × 2 Willis's to do this using your approach. 

 

By way of an example, my SunAmps and my Willis are switched using decent SSRs.  My electrician directly wired the flexible 16A connection cables into the SSRs.  The problem with this approach is that the SSRs run hot (over 60°C when switching 12A) and I had a couple of connectors fail by meltdown, needing me to replace these SSRs.  This happening was a concern.  In the end I addressed this by adding another power connector strip and using a 2.5mm single core link from each SSR to the connector, wiring the flexible phase into the connector as well.  This 2.5mm single core is fine at 60°C and the connector strip was at ambient, so this was fine for the crimped multicore. 

Edited February 16, 2022 by TerryE

[joe90]

11 minutes ago, ProDave said:

If you mean you have a Fused Connection Unit at the immersion heater, best replace it for a 20A double pole switch.

Sorry, just looked, they are 20a double pole switches fused at the board. (My lecky was good).

[ProDave]

10 minutes ago, TerryE said:

The problem with this approach is that the SSRs run hot (over 60°C when switching 12A)

You do have them on a heatsink?

[SteamyTea]

37 minutes ago, Ajn said:

those trades that borrow your extension lead but do not unroll it all and it comes back as one working but stuck together blob.

I left my extension lead, unrolled, when I was working on a sauna.  Popped out to the van, less than 5 minutes, came back to darkness.

Someone had taken the plug off my lead.

Bastards.

[TerryE]

12 minutes ago, ProDave said:

You do have them on a heatsink?

 

Yup they are decent Din-rail mounted Crydom SSRs with heatsink built-in, but the triacs have a 40-50 mΩ resistance meaning they dump ~6W heat at 12A so the contacts still get hot.   The other thing that I am really unhappy about was that my electrician used an ABS case, and I really would prefer a steel one.   I am also considering using these Sonoff SPM switched Power Relays, but that's a different topic. ?

[Radian]

2 hours ago, TerryE said:

 I am also considering using these Sonoff SPM switched Power Relays, but that's a different topic. ?

 

In which case, for mqtt, Tasmota's got you covered  (you probably already knew that because you were going to roll your own in LUA).

[TerryE]

@Radian, no I would use Tasmota and MQTT.  The MCU internally uses an ARM core, but the Tasmota support for this is in Beta ATM.  I particularly like the fact that you can Enet connect it to your RPi or whatever you are using to control it, and this makes it practical to use a steel case.  This pricing seems ridiculously cheap, but it is CE approved and the design and build quality as far as the 240V vs 5^-V separation is better than the SunAmp controller boards.

 

I also like the in-built power monitoring so you can track the power use of your subsystems.

[SteamyTea]

9 hours ago, Radian said:

One additional thought - if installing a pair of heaters for redundancy, consider wiring the two 3kW elements in series. This will obviously reduce the power to 1.5kW

Right, thinking cap on.

I have E7 heating into a 200lt water cylinder. Two 3 kW elements (nominal). One at bottom and one at top of cylinder. You may have seen my charts about the temperature gradient elsewhere.

If I where to wire those two elements in series, and with a bit of experimentation with the two built in thermostats, I could possibly run them off a very cheap PV installation (maybe modules hanging on walls) and a cheap 2 kW inverter (about 40 quid).

The thermostats could probably handle 6 or 7 A DC, but I could test that first, and there would be 2 in series, so already a bit of failsafe built in. Seem to remember that a capacitor, or is it a resistor, can reduce arcing.

What do you think?  

[Radian]

You seem to be talking about the thermostats on the immersion heaters, so why are you concerned about their switching capacity? The current they break will be half of what they're normally dealing with when seeing the full 240V across their loads. But yes, this might be a cheap way to bring the load down to something a relatively small PV could supply. I'd be interested to know where you can get a 2kW invertor for £40. A while ago I bought a second hand ABB 3.6kWp for £50 and thought that was a steal! Picked it up to test some modules out of a PV supplier's liquidated stock but haven't got enough to get past the 120V minimum input voltage. Just taking up space at the back of the garage now.

 

Oh, and capacitor/resistor (snubber) networks are sometimes used to prevent contact arcing - but only for inductive loads. The sudden interruption of current causes the collapsing magnetic field to set up a large voltage as it 'attempts' to keep the current flowing around the freshly opened circuit. A capacitor across the switch contact will present a low impedance path (zero impedance initially) so a low value resistor in series with that will limit the current, absorbing the power stored in the collapsing magnetic field and turning it into heat. Not needed for switching purely resistive loads although there's always parasitic inductance lurking around in the wiring so has to be evaluated on a case by case basis.

[SteamyTea]

9 hours ago, Radian said:

so why are you concerned about their switching capacity

If I did not bother with an inverter and just switched DC.

Edited February 17, 2022 by SteamyTea

[Nickfromwales]

Just to bring this back down to Terra Firma for a mo;

The Willis heaters are built, tested and sold on the basis that they will perform admirably as delivered, eg -3kW max delivery switched by the in built stat.

The reason I use them, and fit 2x, is that you can dump circa 6kWh into a heat load at any one time, reliably. If there was a remit to dump 1 or 1.5kWh linear I would choose a different device. 
 

17 hours ago, Radian said:

It seems that people can be misled by thinking that if a 3kW kettle is OK using a three-pin 13A plug then any 3kW load is similarly safe. The kettle, however, is only operating for seconds and any warming of the lead and plug goes unnoticed as the heat has insufficient time to build up.

This is almost scaremongering afaic….

This thread has discussed Willis heaters, not kettles, so for this thread to be of value, let’s keep the apples and apples together  

For completeness, Willis heaters would absolutely never be on 13a plugs, and always on the 20a DP switches that they require for safe, reliable use.

 

I installed some Willis heaters once, alongside my sparky, and they went in hardwired into 13a fused spurs fed from 16a rcbo’s. I subsequently got shot down ( aka re-educated ) here and was duly informed that they should have been on 20a DP switches ( which my company then instructed my sparky to return to site to change accordingly, with immediate effect ). That made me then go away and revisit this, as more than a few of my new build ‘passive’ clients had also expressed an interest in going “all electric” so I wanted to be sure I had a robust method of going forward with their heating / DHW arrangements. Ergo, I am now VERY familiar with this method and discipline of heating, and its execution. I have been granted access to data collection from some of these all electric clients, so I hope to be able to summarise in a year or so across the board with the different dwelling types.
 

The above-mentioned Willis heaters had, however, been in for a full winter ‘heating’ season, fitted into a PH which used 2x Willis heaters for CH, and none of the system components showed absolutely any evidence of degradation / fatigue etc. The use of bootlace ferrules is of huge significance here, as no known heavy loads on flexible cables should be made off without them, a-la @TerryE’s comments above. 
 

There would only be issues with “boiling” aka kettling if there is insufficient water flow / circulation or the in built stats on the Willis heater(s) we’re set too high. This is all either controllable or can be mitigated by design. 

[Radian]

1 hour ago, Nickfromwales said:

This is almost scaremongering afaic….

This thread has discussed Willis heaters, not kettles, so for this thread to be of value, let’s keep the apples and apples together  

For completeness, Willis heaters would absolutely never be on 13a plugs, and always on the 20a DP switches that they require for safe, reliable use.

 

I fully understand and respect your POV which appears to have been been formed through experience and sound technical advice. However, it cannot be denied that both 3kW electric kettles and 3kW Willis heaters look absolutely identical as an AC load. Yet, as you point out, one is routinely equipped with a 13A plug and the other " would absolutely never be on 13a plugs" - in your own words.

 

To identify and explain the actual reason behind this difference is, I feel, valuable information - not scaremongering. It should certainly not scare your clients as you are doing things correctly. But this kind of 'gotcha' persists throughout the self-build environment and I've seen other members making mistakes along these lines so feel compelled to maintain its visibility wherever I see it arise.

 

 

[TerryE]

@Radian, a kettle is on for no more than the minute or so need to boil a litre or two. A Willis is designed to be on for hours.  These cases differ in time profile and therefore in thermal dynamics.

[SteamyTea]

On 16/02/2022 at 09:06, Radian said:

will reduce the current draw to one suitable for standard 13A plug and time switch

@TerryE

I think the topic of load through a 13A plug came from the above.

I can see the thinking behind it. I can be a quick fix/test if a concept. Probably not a permanent solution.

[TerryE]

@SteamyTea, either way I would be very uncomfortable leaving a 12A load connected through a standard UK plug for many hours overnight unattended. They just aren't designed for this type of use.  Seems a recipe for a fire where the insurers could reject the claim. 

Edited February 17, 2022 by TerryE