Direct to water heating.

curiouscarbon

Science Penguin
Joined
Jun 29, 2020
Messages
2,136
In AC, the current and voltage is measured in what is known as 'Root-means-Squared.' This is a complicated mathematic way of measuring AC current and voltage in 'DC equivalent' numbers when using a resistive load. If you look at a 120V AC signal on an oscilloscope, the voltage peaks are at 170V.

Note Root-Means-Squared can be thought of as the average..... but it is not really an average. The mathematical average of an AC voltage or current is zero.


The net result is that if you have 100V going through a 100-ohm resister it results in 1 amp regardless of AC or DC. It also burns 100W in both AC and DC.

However, When the load is inductive or capacitive, the equivalency starts breaking down.
really silly question, any situations where the mean of the absolute value of the voltage with AC is used? (that won't be zero with AC)
 

rin67630

Solar Addict
Joined
Apr 29, 2020
Messages
873
Dont forget: the regulator elements on the boiler are dimensioned for a given voltage.
And the breaker contacts won't deal well, if the switchin frequency is too fast to let the arc extinguish.
I would not switch faster than 100Hz.
Switching 3KW at 250V is noch much more than ~12A. Peanuts for good FETs.
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
A solar panel is a current source. Charging a capacitor at power point voltage and discharging that into a resistor with PWM for heating should be very efficient if delta V is kept very low. The greater the voltage difference, the higher the losses. Am I correct that this is a good real world example.





Sam Ben-Yaakov

Sam Ben-Yaakov
2 years ago
Yes the losses are in fact liked to delta V. So the smaller they are the higher will be the efficiency.

Just to put this to rest, here is a noted professors opinion. If you want to learn electronics correctly, watch this professors videos.

The FET on this board don't heat up at 10A and it operates at the equivalent of 50 Hz with arc interrupt so heating elements can be daisy chained with common upper thermostats and it can be paralleled with a charge controller's array with no battery needed.

 

rin67630

Solar Addict
Joined
Apr 29, 2020
Messages
873
A solar panel is a current source. Charging a capacitor at power point voltage and discharging that into a resistor with PWM for heating should be very efficient if delta V is kept very low. The greater the voltage difference, the higher the losses. Am I correct that this is a good real world example.
In a practical use, the efficiency is just close to be irrelevant.
You will want to dimension the system to be able to produce warm water with a cloudy weather as well. It will do, but pretty slowly since you have < 20% of the max power available. The heaters will be massively underpowered.
With the same system on a sunny day, your boiler will be a at temperature at 14H and the rest of the time you will be throwing energy away...
All non-grid connected systems have the same problem..
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
I have a rather minimal system and my water heater is at 60C by 10:30. That is with only about 500W maximum going into the heater. All that is just spare power that would normally be wasted. There is always too much power available on good days. They mentioned 30% increase with MPPT, that is true for overall average over direct connect. However, it is in less than ideal sun that a MPPT really performs over direct connect. Small amounts of power really add up.

Panel MPPT Increase over
Current Direct Connect
100% 0%
90% 10%
80% 25%
70% 50%
60% 67%
50% 100%
40% 250%
30% 333%
20% 500%
10% 1100%
 

rin67630

Solar Addict
Joined
Apr 29, 2020
Messages
873
Panel MPPT Increase over
Current Direct Connect
100% 0%
90% 10%
80% 25%
70% 50%
60% 67%
50% 100%
40% 250%
30% 333%
20% 500%
10% 1100%
Can you explain these figures?
They appear frankly unbelievable.
If you have e.g. 100% more at 50% that would mean you always get 100% output irrespective what you put in?
:eek:
 

houseofancients

Photon Sorcerer
Joined
Nov 29, 2019
Messages
1,081
Can you explain these figures?
They appear frankly unbelievable.
If you have e.g. 100% more at 50% that would mean you always get 100% output irrespective what you put in?
:eek:
i suggest you search his yt channel with real life test and figures
 

houseofancients

Photon Sorcerer
Joined
Nov 29, 2019
Messages
1,081
I won't do that. Pardon me, but I hate the trend of discussing everything on videos.
not discussing , by watching his video's you could see that @efficientPV or opera as his channel is called, has been building ,running and testing this for quite a while, instead of just theorizing
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
Can you explain these figures?
They appear frankly unbelievable.
If you have e.g. 100% more at 50% that would mean you always get 100% output irrespective what you put in?
:eek:
This represents the improved performance over a direct connect system.

For example, at 50% rated panel current you should be able to get 50% of the panels rated power with MPPT.
With direct connect, 50% rated current into a fixed resistive heating element the result will be 25% of the panels rated power. Direct connect heating at this power level would require twice as many panels to produce the same heat as a MPPT heater control.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
5,414
Location
Los Gatos CA
I have done *small* direct heating for a battery box warmer and was careful to match the heat element resistance to the Vmp and Imp of the panel. It is still not as good as using an MPPT, but I found it good enough.

 

rin67630

Solar Addict
Joined
Apr 29, 2020
Messages
873
This represents the improved performance over a direct connect system.

For example, at 50% rated panel current you should be able to get 50% of the panels rated power with MPPT.
With direct connect, 50% rated current into a fixed resistive heating element the result will be 25% of the panels rated power. Direct connect heating at this power level would require twice as many panels to produce the same heat as a MPPT heater control.
I knew that these math were wrong !
Direct switching PWM can only simulate a higher resistance, but can't change the voltage.
With oversized panels (in voltage), you may in a limited range adjust the apparent resistance to match the available power.
But if you oversize the panels, the boiler regulator's switch will commute a voltage for which it has never been designed and finally the voltage will always be the limit or you end up with burnt temperature controllers.
 
Last edited:

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
I have done *small* direct heating for a battery box warmer and was careful to match the heat element resistance to the Vmp and Imp of the panel. It is still not as good as using an MPPT, but I found it good enough.

On the days when the sun shines most will likely not be the time the most power is needed. There is a funny fact about the ideal resistance. One study indicated that the ideal resistance could almost be doubled and the total weekly power would not drop that much. This would give more consistent heating throughout the day and much higher production on the poor days.
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
I knew that these math were wrong !
Direct switching PWM can only simulate a higher resistance, but can't change the voltage.
With oversized panels (in voltage), you may in a limited range adjust the apparent resistance to match the available power.
But if you oversize the panels, the boiler regulator's switch will commute a voltage for which it has never been designed and finally the voltage will always be the limit or you end up with burnt temperature controllers.
That is an unrealistic application since a majority of the heating elements are rated at a much higher voltage than the arrays which are common. A slightly lower resistance than normal is used to compensate for thermal effects of the panel. Dramatically lower resistance changes the RC delay characteristics creating poor performance. Interesting to note the ACTii/PL schematic does not show any heater over temp switches and refers installing an additional solid state relay to disconnect in an overheating situation. They rate their device up to 350V.

Here is another German all in one entry to the fray. Good luck with life keeping the capacitors hot all day. It is limited to 45V and 550W though thy suggest possibly 1200W in parallel panels to power it. Costly and not that efficient from the looks of the board.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
5,414
Location
Los Gatos CA
On the days when the sun shines most will likely not be the time the most power is needed. There is a funny fact about the ideal resistance. One study indicated that the ideal resistance could almost be doubled and the total weekly power would not drop that much. This would give more consistent heating throughout the day and much higher production on the poor days.
Interesting. Can you link to info on the study?
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
Interesting. Can you link to info on the study?
Unfortunately, that link was on a laptop that died this year and I don't have easy access to the document without heavy searching. But I do have the relevant chart which shows production in different locations reflecting different weather conditions.
PV PvsR.PNG
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
This is to provide further clarification to the above chart. The specification for the 2,000W array is as follows using eight 250W panels. If you follow the advice of David Poz, the ideal resistance would be 28.4 ohms. Compare where that puts you on the chart.
panel spec.PNG
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
5,414
Location
Los Gatos CA
This is to provide further clarification to the above chart. The specification for the 2,000W array is as follows using eight 250W panels. If you follow the advice of David Poz, the ideal resistance would be 28.4 ohms. Compare where that puts you on the chart.
View attachment 94910
I assume the 8 panels are in series?

For 8S arrangement:
Vpm = 8 x 29.8 = 238.4
Ipm=8.39A
Rpm= 238.4/8.39 = 28.41

It is interesting that the peak production in the chart is between 40 and 50ohms (depending on latitude)
  1. Istanbul: 41deg N
  2. Genova: 44.4deg N
  3. Bordeuax: 44.8deg N
  4. Zurich: 47.4deg N
  5. PraGUE: 50.1deg N
When I calculated for my small box heater, I just aimed for Pmax. Now that you point it out, it kinda makes sense that making the resistance higher will give you a higher average power over the whole day.

Once we accept that average production is more important than peak, the effect of latitude also makes sense. The cities in the studies are all fairly far north. It would be interesting to see what the curve for Miami (25.8deg N) would be. Just doing an eyeball extrapolation, it looks like the ideal resistance would be a lot closer to the 28.4 Rpm value.
 

efficientPV

Solar Addict
Joined
Sep 24, 2019
Messages
604
It may look that way. Latitude has more to do with the weather conditions for these cities, the presence of clouds. It is amazing how much production lowers with only a small wisp of a cloud. I devoted a panel to a fan for the back yard where my wife lounges. You won't notice anything and the fan suddenly drops in speed with almost nothing for a cloud. The eye is log and panels are linear. For a 10A panel, peak current for the day may only be 7A so you better design for it. Power is a square function of the current. Pick your poison. Some focus on getting the max under beat conditions. In peak sun there will always be more than you need. For me, I need to have a system work on the worst day.
 

brum

Solar Enthusiast
Joined
Feb 8, 2022
Messages
96
If you go for a system that works on the worst day there will be plenty of excess energy that you can use for other purposes. In such a case, I would go with a grid-tie inverter, not with direct water heating.
 
Top