Ditching gas, going electric immersion only, a wee project

Solarchaser

QrizB said:

Solarchaser said:

I've been trying like a bear to find a double isolating high temperature replacement thermostat (x4) to run say 90C with a 95C manual reset, but the various stores I've visited and search terns I use have not came up with any usable solutions.

Have you thought of switching from cheap, passive thermostats and moving to a more industrial approach using thermocouples?

This sort of thing (not a specific recommendattion, just the first likely-looking example I could see):
https://www.ebay.co.uk/itm/224040936181

I've run lots of lab- and pilot-scale rigs and controllers like these are pretty foolproof. Depending on how your tank would respond to boiling, if you want belt-and-braces you can run two controllers in series; one set to the working temperature and a second as an overtemperature trip.

After initially dismissing these as a solution,  I find myself in an embarrassing boris Johnson style U turn. 

The solution I wanted to use just is not accurate enough.
Hands up @QrizB and @Heedtheadvice, I was wrong, and you were right.

I had a few weeks where all ran relatively smoothly, but manual heat cut outs have tripped several times in the last week meaning no solar heating. 

The hot block calibration worked pretty well for setting the cycling thermostats, but the manual reset is independent of this and goes by the temperature on the base of the thermostat rather than the rod, and other than trying to insulate it to increase the trip.... which is far too hit and miss, I find myself having to concede defeat with these bloody things.
I just dont trust them.

So onto these little controllers you linked to, I note first off they have a solid end K thermocouple,  so should be more reliable than the ones I used to have to deal with.

My largest issue is that I want them to essentially be on all the time until they cut out, 24/7/365.
And ive had few dealing with solid state relays, but I always remember they don't like being on for lots of time. (Much like manual relays)

So my question is do you get normally closed relays (of course you do) that the controller can switch to open when a set temperature is reached rather than having closed when power is applied.
Or relays solid state or otherwise that are happy to remain on indefinitely without dying

Heedtheadvice

see edit first!!

Only just read your post!

"So my question is do you get normally closed relays (of course you do) that the controller can switch to open when a set temperature is reached rather than having closed when power is applied.
Or relays solid state or otherwise that are happy to remain on indefinitely without dying"

Will consider and post back but a couple of first thoughts...

As you write, yes you do get normally closed relays or even changeover ones so you get NO and NC in the same package.

One thing to consider is the fail safe action. For a NC relay it requires power to open. If power is !ost to it's circuit whilst the remainder of your system is powered (i.e. a local fault) then it would fail in a potentially dangerous or unwanted state not cutting out and stopping what it is supposed to protect! Consider NO could be the better option.

NO relays that are powered all the time should be very reliable, with correct selection to suit circumstances.....but certainly not indefinately but long enough time to fail.

To do a good selection you might post (or pm me) requirements such as:

Switching voltage, normal switching current and that under fault conditions (if you can) ambient temp (max and min). I assume the area is dry but advisable to cope with leaks etc so maybe IP56 or 66 environmental protection or suitably enclosed. Operating voltage and current needs considered but that could be for later!

EDIT: Just had second thoughts. Previous post suggesting thermocouple was regarding temperature control rather than a safety cut out which I think is what you currently refer to? For that you need not just a reliable and repeatable component but a safety 'system'....a simple one. I refer back to the above comment on fail safe. Maybe a ready made cut out ...that suits your purpose....would be the choice. This really depends upon if there is a safety problem if the 'cut out' does fail to operate. You can perhaps answer that first!!

Solarchaser

Thanks for the reply.
Mixed bag really.
When the eddi relay pcb gets here eventually it can take 2x pt1000 sensors which I have, and can set a max temperature within its software at which it will shut off the heaters.
I'd consider this to be pretty good going by my myenergi experiences so far, so I'll trust it to heat 2 of my 4 elements.
Each element is 3kw, individual mcb off 16a and isolator.
So ~12a each.

Though I might combine the two elements not attached to eddi on a 33a mcb instead to utilise the ssr's

The controllers QrizB linked, I've spent a few nights researching those, the solid state relays attached to them also, and to be honest have bought 3. They should be here in a couple of days.
My overall thinking is they are 40a each, so safely 2 elements each.

So 1 to run two elements at say 90C limit, 1 as a backup at 95C incase the first one fails.
1 to run 95C for a failsafe on the eddi.

From what I've read the solid state relays are pretty happy to sit on for ever and a day as long as you are not maxing them out all of the time, and with the extreme case being 24a for say 4 hours, on a 40a relay over a 24 hour period, its certainly not maxing them out.

I agree completely about the default to nc, for exactly the reason you said, a fail causes permanent on rather than a safe fail for no devices,  so the ones I have ordered are resting state of normally open.
It was an idea for less stress on components a couple of nights ago, but you always have to think "what's your worst case scenario" and so it had to be no.

The area they are in is essentially dry but subject to outside temperatures to some extent bearing in mind they will be beside hot water cylinders, so -15 I think was what we had about 7 or 8 years ago maybe.
I'll be putting the controllers in some sort of box, but will likely leave the relays open to air for cooling.

Worst case would be a boil, but really its not a massive issue, at 7bar or 95C the safety's open on both tanks, those are (will be this weekend) plumbed to exit outside the foundations and under the decking, so it will be annoying to have to top up the tanks, but realistically the chances of them boiling away 400+litres of water and me not noticing the steam rising from below the deck, pretty darn slim I'd say.
The chances of me not noticing a current draw outside my "cheap" hours are absolutely zero 😂

Solarchaser

Minor update on the actual heating.

Last night I wired in the double timer to the destratification pump.
So now after the central heating pump has been running continuously for 20 mins, the destratification pump will run for 10 mins.
So if the central heating is pumping continuously for an hour, the destratification pump will come on twice for 10 mins each.
I believe this gives me a more accurate picture of how much heat I am using/losing to heat the house.
With this running the top and bottom tank sensors are now around 2C difference instead of up to 10C difference before the destratification pump was running, so it's doing a pretty decent job of mixing/moving the water about.

This means I have a more accurate picture of how much energy I'm using to heat the house, which is considerably more than I thought previously. Due I assume to the heat being drawn out from only 1 side of the tank and the sensor being in the middle, and now with the water circulating I get the whole tank at the same temperature. 

Previously I believed I was using around 6C an hour, or 3kw an hour to heat the house, however in reality it is a little under double that when heating the house from cold.
Outside temperature this morning was indicated at 6C, and to take and maintain the house at 21C it used 8.5Kwh in 1.5 hours, which is to say it dropped 17C from where it was last night, so I guess some could be heat loss in tank etc as not fully insulated yet (still messing about)

I'm hoping it will use less to maintain a constant temperature, not least because the pump will be kicking on and off to do that, forecast says its to be around 11c after dark, so I guess that will be tested tonight.

On the plus side, due to element trips the tank had been cooler earlier in the week and I've found that even at 45C the tanks do provide heat enough to lift the temperature in the house from around 15C to around 21C Iver the course of around half an hour, so I have more useable range than I believed. 

Patrol

This is all fascinating. What are your thoughts on it being a material factor when the house is sold (will/should the buyer be aware, are they likely to be cold shouldered regarding maintenance or repair, will it have to be removed / certified) or your wife survives you for many years. Am trying to understand if there are downsides that should be factored in, not meant as negative.

markin

What temp do the elements melt/fail, If it wan't for the plastic coil i would say fill the tank with sand, daft idea?

EricMears

markin said:

What temp do the elements melt/fail, If it wan't for the plastic coil i would say fill the tank with sand, daft idea?

Afraid so !  See https://www.education.com/science-fair/article/heat

Solarchaser

Not planning on selling the house, just spent the last 4 years redoing pretty much every room in the house.
The system is so... I dunno, geared to me, that If I was run over by a bus tomorrow,  I'd imagine the wife would get a combi put back in a couple of weeks after,  as I'm doing alot of monitoring and I guess tuning at the moment and it would be fair to say that anyone coming in would struggle to understand how to make it work.

There shouldn't really be any maintenance,  it should all just work.
1 valve, 2 pumps and 4 elements are really the only possible fail points.
Compare that to a combi boiler (I've limped one on for last 8 years, so I know) where alot more maintenance and more complication is involved, but it would also be far more familiar to any normal heating engineer.

But that is true to some extent for my extra solar, and my home batteries,  so I guess the wife will have to wait a few years to poison me 😁

Markin, coils are stainless steel, no idea what the elements will fail at, but I wouldn't fancy trying to remove them in a sand filled tank.

I assume that you are referring to warmstone or tepeo which use something along those lines, but the conversion losses are fairly high, conversion losses for heating water are pretty much zero

Edit. Just got an update that the myenergi eddi relay board I originally ordered in May is delayed again till end of October.  😥

Heedtheadvice

Given your thought that a normal safety cut out omission is acceptable then the only issue would seem to be noticing the venting. Only you can judge both those aspects.

I was going to suggest a capillary based sensor/ switch type of cutout. Not relying on electronics they are inherently reliable especially from abnormal waveforms or power fluctuations etc. They are quite cheap but would need to be well mounted physically to ensure good heat tranfer. Not beyond you I expect.

If any interest the following is typical https://uk.rs-online.com/web/p/thermostatic-switches/7334717 , available in auto reset (wide temp differential to reset) or manual reset. Could also be used as an 'alarm' switch!

Solid state relays are quite reliable and under run as you plan should last well especially if you mount them somewhere cool and can add a good heatsink. I would look at their spec to also research their cold limit. Many electronics are not specced down to -15C but will have military versions which are much wider spec. EDIT: Those I just had a look at seem to be fine cold but high temperature remains a consideration for absolute max op temp and potentially derating with temp rise.

I would take care using one with an Eddi. SSRs are design to work with full wave low harmonic sine waves. If the Eddi does not give such a satisfactory waveform on it's output then that might be an issue. Similarly two SSRs in series could be a problem. No issue if using an Eddi mechanical relay output. If any doubts Myenergi might be able to help? Design notes or application notes readup might be worthwhile for the SSRs! EDIT short comment: electromechanical relay connection quite simple, circuit requirements for SSRs not so and fast acting fusing ought to be employed too

Series electronic controls like you plan would not have been something I would have designed without further research - not saying it would not work but I have my doubts.

Sounds like a great system nearing completion though.

Solarchaser

Thanks very much for your insight.

In terms of the eddi, I was actually planning to run the eddi power through the ssr, so that if temperature is exceded, power to the eddi is removed, my thinking there was safety, but also if the eddi is not controlling the power right, id assume it would be a software issue, so a reboot would help.

The capillary sensor amused me, its one if those I use to control the radiator fans on my track car 😁

I think temperature wise I *should* be OK, the "circuit" will be around half a meter away from the tanks themselves (electrics and water get on like a house on fire) so shouldn't get too hot, and I have a couple of pretty chunky heat sinks that would normally keep lasers cool, so should be good with that side of it, there are fans on the heatsinks which I can use if I feel they are getting anymore than slightly warm.

I must admit I didn't consider the series circuit to pose an issue, because although I know they are solid state and so semi conductors, I was still thinking of them in terms of physical relays so clunking on one strap rather than a whack of transistors switching over, and so I can see the potential for issues if one relay is switching on and off with a borderline temperature and giving the other relay a bother.

Hmm, I suppose the pragmatic approach at this point is to suck it and see.
I'm pretty confident the eddi power one will be fine, and if the other one is problematic, then I'll switch to a mechanical fail safe I guess.

With the weather turning now and this relay board delayed another month I have a decision to make about whether to just run the system on 3 elements, hoping that we don't have a proper cold snap, or whether to leave 1 element only on eddi and power the other 3.
A wee ponder required.