 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Solar : Poor mans PV Diverter.
Page 1 of 2   |
|||||
| Author | Message | ||||
| George65 Guru  Joined: 18/09/2017 Location: AustraliaPosts: 308 |
I have come up with an idea for a simple, cheap and cheerful if not perfect idea for a Solar PV inverter to send power to my HWS. Seems the trick is to work out when the solar is generating and divert that power rather than something like a timer that cannot work out if there is solar generation or not and will use the expensive grid power on cloudy days. I have seen lots of complicated/ expensive soloutions that I have no doubt would work well but are over my head to build or are way too expensive for what they are to buy. Waiting for the chinese to build a board and start selling it on fleabay for 12 Bux but in the mean time, this is what I have come up with and the idea I offer for Sacrifice and public Stoning..... When the Inverter is back feeding into the line, the voltage goes up. The more input the higher the line voltage. My idea is to use this increase in line voltage to signal there is solar generation and to use it to switch the HWS. The component I plan to use is this:  A voltage monitoring Relay. Once the solar starts backfeeding and increases the line voltage to the prescribed level ( yes, some tuning will be required) it switches in the HWS. The relays appear to have an on and off voltage hysterisis so the drop because of the load should be accountable for. If a cloud comes over or output drops below threshold, the relay should kick out till the voltage comes back up again. I plan to have the voltage monitor feeding a higher power relay which in turn will feed a PWM controller that goes to the heater element so I can control the amount of power I want to put in the heater. As I will likley make up the control unit box with these components, breaker etc ( and anything else that may be suggested) I'm thinking of doing it as a plug in to an outlet near the heater. As my heater is 4.8Kw and the outlet isn't going to wear that ( nor will my solar generation input) I'll turn the PWM down to say 2000W. On a cloudy day when there may be no solar generation to speak of and the cheapest power will be the .11C from the off peak, the heater won't get any boost through the day and will function as normal. When it's bright and sunny, the offpeak won't need to kick in or will just make up a bit for any shower taken in the evening. No, it's not a perfect solution and I don't thank many are but it's cheap which means will pay for itself quickly, easy to just screw together for the electronically ignorant ( like me) and I think should work well for what it is. This is a link to the voltage monitor I am planning to use: Voltage monitor I'll now go put on my crash helmet and hide under the table while everyone shoots the idea down in flames and points out what I have overlooked. :0) |
||||
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
You would be far better off with a current sensing device as the grid voltage could go up and down because of other factors than just your PV. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
Can you recommend a simple, easy to implement devise of this type? |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Solarmiser is one option, I am sure there are others. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Mulver Senior Member  Joined: 27/02/2017 Location: AustraliaPosts: 160 |
George. I agree with Madness. These are cheap as chips to make. https://learn.openenergymonitor.org/electricity-monitoring/ctac/how-to-build-an-arduino-energy-monitor Then you know exactly what's going on and then turn things on and off accordingly. |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
Can anyone see inherent problems with the Voltage monitoring relay? Line voltage is pretty stable here in my observation and certainly doesn't climb to the level of when the inverters are back feeding. As far as I can see, using the voltage rise to trigger a relay should satisfy pretty much all the parameters. Not looking to split the current between home and hot water, I can effectively do that with a PWM set to do 2KW or whatever. Just ordered a couple rated at 6KW so should do the job. The solar miser is more than I want to pay just yet, unless till I try the cheaper option with the relays and see how that works and what features I may want and need in a more elaborate option. The open energy monitor looks interesting and would certainly have a lot of learning value it it for me. Something I'll spend some time on and start getting my head around. Unless anyone can point out some major flaws with the voltage monitor, I think I'll give it a go. Has to be a step up on a timer and seems the most simple, cheap and effective way of doing it I have seen so far.  |
||||
| Alastair Senior Member Joined: 03/04/2017 Location: AustraliaPosts: 161 |
Before committing to using the voltage switch which looks good for the price, i would do some simple tests. You may have already done so. Rig up a simple switch to turn on the hw and watch the voltage changes with a meter as other loads or sun variations occur. The device you have found has adjustable levels, hysteresis and time delay which is pretty impressive for the price so I would think you should be able to make it work. sorry to state the obvious when I reread. Cheers, Alastair |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
I have a reasonable idea of the voltage changes with the loads. Haven't observed the hot water obviously but I have seen the AC many times. That only pulls the line down 3-4 Volt which is good because the it's a lot less than the inverters drive the voltage up. That can easily be 10V on a good day. The only other thing that really has an effect on the line voltage while the inverters are running is the electric stove and oven. When the Mrs has both on, that can sag the volts a bit. Don't see it as a problem though. She does some cooking, hot water kicks out and that's it. Even if it does not kick in all day because she's cooking up a storm, I'm twice as well off paying for the water to heat up at the off peak rate of 11C kwh than running the stove at .30C. I think while it's still summer-ish, I'll pull the off peak all together and see how we go just running on solar. Should have a few days capacity in the tank anyway. I intend to run some bigger cable from the shed probably right back to the board so I can pump a heap of power back down and that will be a separate circuit I can tap with the voltage monitor and go to the hot water. I'm encouraged the educated people here can't see more wrong with this. I know it's not perfect but for the money which would cost me less than one quarters hot water bill as against the next cheapest solution's which would be over a years return on my hot water costs, I think it might be best bang for the buck. I doubt it will eliminate my hot water costs but i'm confident it sure will put a big dent in them especially when I get the other 7KW of panels I have waiting on the roof for a total of 14. Might go to 20kw for good measure.  I'm sure it will take some tuning but then what I'm reading of the more elaborate solution's, so do they.  |
||||
| Alastair Senior Member Joined: 03/04/2017 Location: AustraliaPosts: 161 |
Seems that the changes that you have are probably large enough to give you the voltage based switching you need. As you say provided it gets most of it right then that is where you are getting most of your savings. Cheers, Alastair |
||||
Clockmanfr Guru Joined: 23/10/2015 Location: FrancePosts: 429 |
Hi George, This looks like a good simple robust and cost effective project. For several years now I have been searching for a PV diverter that is simple, reliable and does not cost silly money. The main reason is to point or recommend a product that has good QC and simple that builders of the OzInverter can incorporate into their own particular system that could be anywhere in the World. I am constantly asked for a circuit/gizmo/unit that can do what you are asking. I am interested to see how you get on with that Aliexpress unit. I will get a few and give them a good test. I especially like the hysteresis alteration/settings as many PV systems have differing changeability's. The ones I have already tested seem to be a bit wishy-washy/woolly in the exact times of hysteresis. The Solarmiser that Madness mentions also looks good for a commercially produced and tested unit and not a bad price. Importantly there is not so much an issue with quality control or setting alterations. Some other units I have tested have hideous programming parameters that just are not sensible for Mr average man. Or just far to expensive. Arduino stuff , sorry lads, but its just to big a step for the majority of The OzInverter builders. And recently I have found some of the Chinese clones to be poor QC and failing after a short period. So George, carry on and I will watch your evolution, thanks. Everything is possible, just give me time. 3 HughP's 3.7m Wind T's (14 years). 5kW PV on 3 Trackers, (10 yrs). 21kW PV AC coupled SH GTI's. OzInverter created Grid. 1300ah 48v. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
George another way you could heat your hot water is just have panels dedicated to heating your hotwater. Only other hardware you need is a solid state relay and a small dc power pack. You need these parts 400Volt Dc 20A Solid State Relay 10 30 VMP solar panels Dc plug pack The SSR needs to able to handle at least 400VDC and 20A, you need this as the thermostat contacts will be burnt to nothing first time it switches off. Instead of switching the panels off with the thermostat we use the SSR just connect the power supply via the thermostat and to the input side of the SSR. Then when the thermostat is closed the SSR will be on, once the HW is up to temperature i will turn off. Will cost you $10 plus the panels. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
Very encouraging to read your encouragement Clockman. This seems a simple solution to a problem as you say seems to have overly complicated solution's. I know there are a lot of geniuses here but I am not one of them and there are unfortunately too many others like me. I had a look at my inverter for todays data and found the low to high voltage was was 234 to 259. There is a lot in that to play with. I saw when the inverter was kicking out 4.8KW the voltage was at the inverter 258. With some cloud it dropped into the 240's. That is at the inverter which thanks to the less than brilliant wireing is higher at that point than at the house. I call this Voltage compression. It's like voltage drop in reverse. Normally with a lead the voltage is high at the mains end and low at the load end. With resistance in the wiring, backfeeding goes opposite with the inverter seeing higher voltage than the mains end. Even if the inverter was closer to the mains wiring as many will be, it still has to push the voltage up to back feed and there should easy be enough in different applications to work from. This is what I am counting on. I don't think the voltage rise is quite linear but it's going to be easy enough to set the voltage monitor at 250 and know the solar is generating when the line voltage is there. If I set the cutout at around 240 ( nominally 230 here) then I can be pretty sure I am in the solar generation window. I have drawn up a layout of the box I want to build to contain the voltage monitor, relay, PWM controller, Volt/ amp gauge, Indicator light for when the power is active, breaker and a manual over ride. Can't wait till the voltage monitors arrive. |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
Mad, That's a great concept! I thought about a similar approach of a dedicated system and bought a couple of DC relays but it bothers me with it's inefficiency. Once the water is up to temp, the generation from the panels is basically wasted. I'm buying panels cheap as chips, bought 16 185's last weekend for $250 but i'd like use the panels generation all the time and this method allows me to do that. This way, although more expensive and less simple than your concept, allows all the power the panels generate to be put to use. Of course I am surmising that there will be excess power in the first place. I have a 400L tank. To get that amount of water up just 20oC with a 1500W input will take a tad over 6 hours. I'd be lucky to get 1500W from 10 185W panels for more than a few hours and the the fall off in winter would make the time/ power factor numbers fall a long way short. having the off peak still operating would be essential in not eventually running out of hot water. My approach allows most of the solar generation ( within other factors) a home may have to be put to use and with a lot of places now having 3-5 Kw and more, it could meet all the hot water needs and without having to fine more room for panels and install them etc. Allright for us nutters but I keep people like my brother's in law in mind and what the make or break for them would be. This goes a bit deeper as well in catering to an ego trip. I'd like to test this out so I can put it out there as a way other people could save money on their hot water and be able to afford/ justify this over other systems and be simple enough for many people to be able to do where something arduino or more electronic would be out of their relm. This would also be applicable to people with " proper" solar setups that just backfeed their excess for some piddly little amount through the day when they are generating excess power. Although some people do better, I know a lot of backfeed rates are around 8c KWh and off peak is at least 11c so it makes sense for most people to put that excess power to use where they can. The way I am looking at this, it's half plugin so could work within what people already have. I think your concept is brilliant and I couldn't quite work it out the direct approach myself with how to do the switching but your concept is unsurprisingly very obvious and ingenious in it's simplicity. Something to keep in mind and adapt to other problems for sure! |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
What you want to do could be done easily and for little money with an Arduino. There may be a cheap off the shelf solution out there but I have not found one. For people who are not technically able to do that the Solamiser is not that expensive and would pay for itself within a year. If that is too much money then start learning about Arduino's. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
This is definitely something I need and want to do! |
||||
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Interestingly enough George, you have identified the things that you can use for a very interesting experiment. You are mentioning that there is a significant difference in the inverter end of the cable compared to the house end of the line. This line loss gives you the perfect way to measure the current direction, and the intensity of the current coming from the inverter. If we think about this, we can put a AC SSR in series with the heater ON the house side of the circuit. If we measure the voltage differential between the inverter and the house, we can get a feel for the power the grid tie is putting out. We can do this by feel or measure ( or deduce) the resistance of your cable, and so W = V^2 / R. We measure the voltage differential over the cable as V, and we know the resistance.. so we can deduce the watts being transferred over the cable. ( we find the cable resistance by measuring the voltage drop for a measured current ( clamp meter?)..so Resistance = voltage drop / current... so now we know what the Resistance is) Assuming the resistance stays fairly reliable over a small temp range, we can get a feel for how many volts needs to be dropped over the cable before we pull on the water system heater. This will be a very simple voltage switch, that will only pull on the water system if there is enough power to run the heater over and above the amount to run it, and not when it is feeding the house, but not enough to run the heater outright. That way we can feed the grid,.... and if enough energy is being generated, we turn on the heater. The act of turning on the heater will not change the differential in the cable, as it is only measuring the voltage loss in the cable, which is purely a response to the current it is carrying, not the house voltage, as the inverter will come down to the house voltage as the heater comes on on the house circuit.... so both will lower, but the current in the cable will be the same as will be the voltage drop we are interested in. This will incur the least interference to the overall system configuration, it will simply turn on the heater when there is enough power coming from the inverter to run it "stand alone" so to speak. It will still be wired into the grid, your just going to turn it on and off when you have the power to do it. You will need 2 very small transformers to supply the reference voltage from the two ends of the cable, a comparator, and a few components... maybe 10 bucks or so. Ideally, a nano arduino would be much better, as you can program the events much better to rid ourselves of too many on off cycles... if that causes you any worry. Or you could modulate the hot water heater element so as to take advantage of all the generated power by controlling the element power differentially via the arduino... ie proportional pwm given the voltage drop... all the way down to very small voltage drop. That way you can get the most from the panel array, by using all the power all the time. If the hot water is already hot, it will shut it's own thermostat, and it will back feed the house anyway... win win I think, as it will use the house as a very efficient battery bank, and diversion load as well. The arduino still needs the two tiny transformers, but then only needs some rectifiers ( tiny) and an solid state triac style relay ... only $4for 40 amp ones.... and not much else, and the nano itself (<$2). I don't like the thought of DC relays, as the thermostat can't switch off safely... particularly if the SSR fails On... which it will. So too will the triac AC style, but that won't matter in this setup, the worst that will happen with failure, is feed the house back into the grid, and the thermostat will function as if nothing has happened, as it sees no part of the system. Edit: Actually thinking further, if you use the arduino nano etc, you could also monitor the water temp, and have a set point (say 45C) whereupon, if it gets that low because of poor solar incidence, you can just switch in the mains regardless of solar inputs, and push it back to say 47C. This is probably enough to have a decent shower etc. This would make it completely autonomous, and take advantage of any solar available, and if not enough to keep it at a reasonable temp, wold do that too. It really is a proportional setup, where you drive the thing proportionally to the GTI input. Should be seamless in operation, and require no input from you once set up. If you don't want the GTI to put more into the house, then you could limit the solar side of the GTI to only generate up to whatever your putting into the hot water (mine or Mads controller will do this). ie limit it to 2.4kw if water is full on, and shut it off if the thermostat shuts the HW off. This will stop back feed into the meter. We can measure this with the voltage drop remember, so the information is already in the nano's grasp anyway. ..........oztules Village idiot...or... just another hack out of his depth |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
How about just having a rectifier connected to both ends of the AC active cable from the inverter that is further away from the house wiring. This will rectify the difference if it is more than 3V it will turn on the SSR.  There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Very simple and elegant, but will not take advantage of generation below 3v ( as an example). Also not sure how the SSR copes with the iffy part of the led almost getting going. My experience is that they are not too good with controlling power linearly, but respond well to pwm. .......oztules Village idiot...or... just another hack out of his depth |
||||
| George65 Guru Joined: 18/09/2017 Location: AustraliaPosts: 308 |
Oz, I would be interested in seeing a layout showing components and a sketch of what you propose. Arduino was the first way I thought about this but couldn't find info I could follow. I have got plenty of controllers, regular and Nano ( was just having a play with them yesterday) and I have got a few of those 30A voltage/ current monitors. I am an arduino dilettante but can follow layouts showing components and connections. I particularly like the proportional control element of your idea. Another thing I REALLY want to figure out is a SIMPLE way to feed power from a a GTI to one leg of My AC but not supply more than the AC is using on that leg or it will try to back feed my electronic meter. A proportional control could be adapted to both applications which would be perfect. Again, would have to have the layouts and sketch an electronic idiot could follow. |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Because in the hot water system, we had a good voltage drop to work with, which made direction very simple... even though we didn't need to know after looking at it, we were really interested in voltage drop  In the three phase leg problem, we have a very different circumstance where the system probably has a 100 amp service behind it.. so a resistive drop is out of the question. Now we have to measure the phase of the voltage and compare them.. I think we XOR them.. ie when we have two zero's or ones, we get zero out, and when we have zero and one or one and zero we get one out. We can square the sine wave samples up with a comparator set for zero crossing for both the current and voltage... and then XOR them. If they are in phase we get zero, if out of phase we get one... so logic high or low.... unless we have a power factor involved, then we must intergrate it, and yet another comparator set such that we get a high or low signal from this according to the amplitude, and that comparator will have the reference set to half the logic state voltage. Thats how I remember seeing it before, and if memory serves me correctly I have Warpspeed to thank for that, as he fully understands this , and I'm..... flaky on it. Long and short is yes it is certainly possible to do all you want, but will require some work from you to construct the circuit boards.... which I suspect to be simple really. Warpspeed is really the man for the phase circuit. Mad has the boards running already for proportional GTI operation, as do I, but mine are at present only for galvanically isolated GTI's, Mads do both types. You will have the non isolated types there I suspect. On the arduino side of it ... well there are many folks here much better than me ( see my GTI control so see how adumbrate I am on that.). I make things work, but the coding is generally not pretty. I will try to scribble out a readable pic of the how to cobble it together soon. .........oztules Village idiot...or... just another hack out of his depth |
||||
| Page 1 of 2 |
|||||
 Print this page |
