DC immersion heater direct from PV

[Chewbacka]

Someone mentioned earlier about measuring the resistance of the heater.  You will get ‘interesting’ results as the resistance of a cold heater element is lower than when it is hot.  On a non optimal day so panels at say 70% this will present more of a load to the panels which will pull the panel voltage down a fair bit, perversely giving less power into the water than you expected.

[Tom Richmond]

27 minutes ago, WotEver said:

No you don’t, because the voltage source (PV) is current limited.

Yes he is. You have a theoretical maximum power source of (4x260) 1040 Watts. Resistive heaters are direct converters of power to heat. A 1kW heater will convert 1kW of electrical power to heat. Whether that’s 21A @ 48V or 10A at 25V it’s still only 1kW

Or 16A at 60v?

Sounds like you guys think my plan will work... And the suggestion that a 48v heating element will burn out if supplied with 60v (or more) seems to not be the case?

So, subsequent question is what sort of thermostat do I fit to the cylinder? I assume I need something with a temperature probe inserted into the insulation, connected to a relay to shut off the power from the panels when tank reaches full temp. How high can I go? 80deg? 90deg? 99deg? Cauliflower has a thermostatic mixer on so hotter the better, as long as I am not boiling inside the tank, right?

[WotEver]

9 minutes ago, Tom Richmond said:

Or 16A at 60v?

Yes. Or 8A @120V...

 

38 minutes ago, Chewbacka said:

Someone mentioned earlier about measuring the resistance of the heater.  You will get ‘interesting’ results as the resistance of a cold heater element is lower than when it is hot.  On a non optimal day so panels at say 70% this will present more of a load to the panels which will pull the panel voltage down a fair bit, perversely giving less power into the water than you expected.

I thought that when it was mentioned but there was already too much confusion and poor advice and I didn’t want to add to the confusion. 

[Flyboy]

25 minutes ago, Tom Richmond said:

Or 16A at 60v?

Sounds like you guys think my plan will work... And the suggestion that a 48v heating element will burn out if supplied with 60v (or more) seems to not be the case?

So, subsequent question is what sort of thermostat do I fit to the cylinder? I assume I need something with a temperature probe inserted into the insulation, connected to a relay to shut off the power from the panels when tank reaches full temp. How high can I go? 80deg? 90deg? 99deg? Cauliflower has a thermostatic mixer on so hotter the better, as long as I am not boiling inside the tank, right?

Doesn't your heating immersion element have a built in thermostat then ? 

[nicknorman]

1 hour ago, Tom Richmond said:

oops. sorry nicknorman. I wrote about the morco, but ended up editing it out as unnecessary clutter. 

So, are you saying that running a PV array of 72voc/60vmp into a 48v DC heating element would not burn out the element? This is the key question for me... 

Yes. The 72v is the no load voltage, the 60v is when the panel is loaded optimally. So let’s say the panels are 1kw at 60v, that is 16.66A. But a 48v 1kw heater has a resistance of about 2.3 ohms. If you could put 60v across that it would be 26A, which is 1.5kw or so. The panels can’t produce that much current.

So in reality assuming a max panel current of perhaps 18A (current doesn’t go up much when you pull a panel’s voltage down below Vmp) you would end up with perhaps 41v into the heater and a power of 740w.

In summary, provided the heater is rated at or above the panel max output there won’t be a problem. But if the panel Vmp isn’t matched to the voltage rating of the heater, some energy will be lost.

As an alternative, you could consider a 110v 3kw heater if such a thing is available here. That will have a resistance of 4 ohms and thus be a closer match to the panels since 60v will push 15A into the heater, much closer to the max power condition of 60v and 16.6A. Note that even though the heater is rated at 3kw, you will only ever get a little less than 1kw going into it.

1 hour ago, Chewbacka said:

Someone mentioned earlier about measuring the resistance of the heater.  You will get ‘interesting’ results as the resistance of a cold heater element is lower than when it is hot.  On a non optimal day so panels at say 70% this will present more of a load to the panels which will pull the panel voltage down a fair bit, perversely giving less power into the water than you expected.

Not by much though. It’s not like a light bulb.

[Tom Richmond]

41 minutes ago, Flyboy said:

Doesn't your heating immersion element have a built in thermostat then ? 

As far as I am aware the DC elements do not have thermostats, no.

34 minutes ago, nicknorman said:

Yes. The 72v is the no load voltage, the 60v is when the panel is loaded optimally. So let’s say the panels are 1kw at 60v, that is 16.66A. But a 48v 1kw heater has a resistance of about 2.3 ohms. If you could put 60v across that it would be 26A, which is 1.5kw or so. The panels can’t produce that much current.

So in reality assuming a max panel current of perhaps 18A (current doesn’t go up much when you pull a panel’s voltage down below Vmp) you would end up with perhaps 41v into the heater and a power of 740w.

In summary, provided the heater is rated at or above the panel max output there won’t be a problem. But if the panel Vmp isn’t matched to the voltage rating of the heater, some energy will be lost.

As an alternative, you could consider a 110v 3kw heater if such a thing is available here. That will have a resistance of 4 ohms and thus be a closer match to the panels since 60v will push 15A into the heater, much closer to the max power condition of 60v and 16.6A. Note that even though the heater is rated at 3kw, you will only ever get a little less than 1kw going into it.

Not by much though. It’s not like a light bulb.

What's gonna be best? A quick Google has not pulled up any 110v immesion elements, whereas 48v 1kw ones are all over the place...

[WotEver]

4 minutes ago, Tom Richmond said:

What's gonna be best? A quick Google has not pulled up any 110v immesion elements, whereas 48v 1kw ones are all over the place...

A 230V 1kW element (widely available) will have a resistance of about 4 ohms so will be better than a 48V one. Probably cheaper too. 

[Tom Richmond]

4 minutes ago, WotEver said:

A 230V 1kW element (widely available) will have a resistance of about 4 ohms so will be better than a 48V one. Probably cheaper too. 

Got one of those already installed. 

So you are saying I can connect the DC output of my solar panels directly to the already existing 230v immersion heater and it will just work? What about the thermostat, is that built to take the increased current that will flow due to the lower voltage? Is the element itself built for the increased current?

Edited April 26, 2018 by Tom Richmond
Edited to add question!

[nicknorman]

12 minutes ago, WotEver said:

A 230V 1kW element (widely available) will have a resistance of about 4 ohms so will be better than a 48V one. Probably cheaper too. 

No!!! Check your maths! A 230v 1kw element takes 4A. R= V/I hence R is 230/4 or 57 ohms.

11 minutes ago, Tom Richmond said:

Got one of those already installed. 

So you are saying I can connect the DC output of my solar panels directly to the already existing 230v immersion heater and it will just work? What about the thermostat, is that built to take the increased current that will flow due to the lower voltage? Is the element itself built for the increased current?

No. Well it will barely get warm anyway. 60V into that will give about 1A hence 60w.

As to the general point, the power rating and the resistance of the elements are fixed (notwithstanding Chewy’s earlier point) and so provided the power rating isn’t exceeded, the current rating can’t be exceeded either. All immersion installations have a thermostat, most have the stat built into the element but of course you need to make sure the one you get does.

[nicknorman]

26 minutes ago, Tom Richmond said:

As far as I am aware the DC elements do not have thermostats, no.

What's gonna be best? A quick Google has not pulled up any 110v immesion elements, whereas 48v 1kw ones are all over the place...

A 48v 1kw one will be adequate. A 110v 3kw one would be better, but not massively so. But if it has a built in stat then that is an another advantage. Anyway, you definitely want a stat or you risk boiling the calorifier water. Not to mention scalding yourself when you turn the tap on.

you might find an adequately rated external tank stat but generally these are not designed to switch the heating load directly.

Edited April 26, 2018 by nicknorman

[Tom Richmond]

1 minute ago, nicknorman said:

A 48v 1kw one will be adequate. A 110v 3kw one would be better, but not massively so. But if it has a built in stat then that is an another advantage. Anyway, you definitely want a stat or you risk boiling the calorifier water. Not to mention scalding yourself when you turn the tap on.

TMV on the output will remove the risk of scalding, but don't want to boil the water inside. 

My thought is a thermostat with a probe inserted through the insulation to the cylinder, connected to a relay switching the DC line from panels to heating element. 

[WotEver]

12 minutes ago, nicknorman said:

No!!! Check your maths! A 230v 1kw element takes 4A. R= V/I hence R is 230/4 or 57 ohms.

Oh yeah, 4A not 4 ohms

[nicknorman]

5 minutes ago, Tom Richmond said:

TMV on the output will remove the risk of scalding, but don't want to boil the water inside. 

My thought is a thermostat with a probe inserted through the insulation to the cylinder, connected to a relay switching the DC line from panels to heating element. 

You can get cylinder stats such as this fairly cheaply (you cut away a section of foam to install): https://www.screwfix.com/p/drayton-hts3-cylinder-stat/95630 But as I said, insufficient current rating of the switch, so as you said, you would need to use a relay as well, to switch the dc power.

It’s still a simpler system though. I think most mains immersions have a duplex system ie a thermostat, plus a thermal fuse type thing in series that permanently cuts the power if the temperature exceeds the max thermostat setting. If you go with a simplex system, I’d want to be sure that my PRV wasn’t clogged up and couldn’t jet boiling water/steam into anyone.

[ditchcrawler]

1 hour ago, nicknorman said:

A 48v 1kw one will be adequate. A 110v 3kw one would be better, but not massively so. But if it has a built in stat then that is an another advantage. Anyway, you definitely want a stat or you risk boiling the calorifier water. Not to mention scalding yourself when you turn the tap on.

you might find an adequately rated external tank stat but generally these are not designed to switch the heating load directly.

Do you think a 1Kw element relying on spare solar will ever get that hot if the OP is using some water each day and it wont be on until his batteries are fully charged. maybe if the boat was left for a couple of days it would be a risk with no draw off and the electrics not depleted overnight. I think I would try it without a stat and monitor how it performs in real life, I don't think it will get that hot

[nicknorman]

7 minutes ago, ditchcrawler said:

Do you think a 1Kw element relying on spare solar will ever get that hot if the OP is using some water each day and it wont be on until his batteries are fully charged. maybe if the boat was left for a couple of days it would be a risk with no draw off and the electrics not depleted overnight. I think I would try it without a stat and monitor how it performs in real life, I don't think it will get that hot

Firstly, you have to look at the worst case scenario and leaving the boat for a couple of days isn’t “worst case” really! It could be fairly routine and you don’t want to come back to find steam and boiling water has been spraying everywhere especially if your PRV doesn't vent overboard. Even if it does, PRVs can stick and that would cause an explosion.

But with the OP saying the heating would start before lunch, you have perhaps another 5 or 6 hours of strong sunshine and a few more of weak sunshine. To get 55l of water from a tepid 30C to boiling using 1kw takes 4.5 hours or so, less time than the potential sunshine available. And if you hadn’t used much water since the previous day, obviously it would be a far shorter time.