Anotherpower.com Forum
Project Journals => Users Projects => Topic started by: dochubert on August 02, 2019, 07:20:41 pm
-
In planning my off grid existence, I quickly realized that water heating was going to be a large challenge. The house came with two 50gal electric water heaters. One seemed to be solely dedicated to the kitchen sink and dishwasher with the other doing everything else.
Running one was bad enough energy-wise, but two was unacceptable.
Redesigning and re-piping followed with the kitchen waterheater removed and the whole house worked off one unit. I did add a hot water return line and a small circ pump so the kitchen had hot water quick.
For off grid water heating the initial plan was a 24v bank of four 200ah batteries running my oldest powerjack, my 24v/8kw. These batteries are separate from my main bank for powering the house. I put in a manual transfer switch so I could select inverter or grid to power the water heater.
I guess I didn't really grasp how much power an electric water heater draws. Switched to inverter and when the water heater tried to run the pj went immediately into overload and shut down. 4500w elements were just too much for an 8000w pj. Plan B was to hook up just 110v instead of 220v from the inverter. Now it ran but of course only heated half as fast with about 2500w measured by the powerjack meter. Also I only had half the solar hooked to it so far that I had planned, so after 2 days use I had to switch the waterheater back to grid to let the batteries come back up. So solar water heating 2 days of each 3. It was a start.
Next, I moved that extra water heater next to the main one and piped it in as a preheat tank for the main tank. Planned to heat this wh with dump load controller from the main battery bank. I installed 2 double 1000w 48v elements into the preheat tank. Got one of these;
[attach=1]
Wired it up to power one 1kw upper element and one 1kw element lower. Planned to get a second controller and hook to the other two elements after trying this one out. It seemed to work fine. The tank heated up nicely, using excess solar the batteries didn't need. Then ran the circ pump to heat the whole loop up and keep the main tank hot enough to not turn on its elements. Eventually, the circ pump will be on a timer and temp switch control.
That all worked well enough for about a month. Then that solenoid relay cooked. I bought it because it was 48v and rated for 440 amps. Figured that was exagerrated but should still be enough to handle the 30-45 amp load I needed handled. Only about 10% of rated. So much for optimism.
While messing with the solenoid while it was failing, I realized the 48v elements were frying too. What the heck? The upper element had opened and was doing nothing, and the lower was cooking. One connection started glowing red hot! More junk overrated parts apparently.
Back to the drawing board.
Time to rethink this whole deal.
After much thought, I decided to put two 2000w 120v elements in place of the 48v elements. Ordered a dc controlled ac relay. I now plan to control the new relay with the same style voltage controller as was used with the bad solenoid. The relay will switch regular ac 120v from my main powerjack as an ac dump load into the water heater elements. Switching ac is much easier on contacts than switching dc.
As of today, I'm still waiting on delivery of the new relay.
Hopefully, this will solve the water heating issues but now I need to think about a real dump load. I want protection for my batteries that I can depend on. Eventually, I will have 3 times the solar that I'm using now. Dump load becomes a must! The water heater setup described above should usually use up the excess solar, but if the inverter isn't working for some reason or the tank is up to temp, ac dump load doesn't dump. So in addition to ac dump, will add a dc dump set a trifle higher voltage using this guy;
[attach=2]
[attach=3]
Its a dc contactor rated 500 amps! This one looks like its actually got a chance of meeting its specs. 60vdc contact rating and a 60vdc coil. I won't be doing anywhere near that of course. I like to use a space heater for my dump load. The heater that looks like an old style steam radiator except on wheels. I pick them up at flea markets and garage sales for $5. They are oll filled and don't have a fan. You need to bypass the thermostat because it can't handle the dc. I just remove the slide connectors from the thermostat and use a large size 30a fuse. The tabs on the fuse match the slide connectors. Connect the slide connectors to the fuse instead of the thermostat. This also gives some short ckt protection. The heater's switches let me use one or both elements. The switches don't much like the dc either, but work ok.(probably shortened lifetime) With one element engaged I see about 5 amps at 55vdc. Different brands of heater are slightly different.
So for cold weather I'll have a heater or two in the basement or back room. In summer I'll probably put some regular dump resistors outside on the chimney out of hands reach.
Back to water heating; (What! Me ramble?)
The 120v 2000w elements are installed and wired and ready. The controller is installed and ready. Just waiting for the ac relay to arrive. Supposed to hit 100 deg F today so hot water without the preheat tank won't be a problem.
That's where it stands. I'll report progress when I have some
-
Hi Doc, I am not clear on why you need dump loads. Surely your regulators will open circuit the solar panels when your batteries are full, or when the water heater thermostat opens.
Maybe with that much power you could investigate hydronic heating. Saves lots of cords and individual heaters running on DC.
Anyway it sounds like you are having fun
Cheers
Pete
-
Hi Pete,
Always having fun!
We mostly heat with wood by preference. That said, I intend to do some amount of warming inside by electric heat thru solar. Then burn less wood to reach comfort level.
The dump load is not truly needed by a solar power system, true. ! like Xantrex C40 controllers. They work very well. I have found that despite settings, if there is lots of sun and no load, that voltage tends to creep up. Batteries are probably the most expensive part of my system so I like to be sure they are protected. The water heating dump will be set to voltage low enough that it comes on and heats the tank well before the charge controllers go to float. The dc dump in winter will be set higher than the water heater dump but not so high that it doesn't come on and move some solar energy into the house for heat. And it will work more and put in more heat after the tank is up to temp.
I do plan to have 3 times the solar wattage I have now when the panels are installed on the roof (soon hopefully). In winter that will hopefully be enough to run the house fairly normally. In summer it will be way way plenty (with the exception of the water heater setup which will be hopefully adequate), so some protection makes me feel better. Two dc dump loads are in my plans, one set to put heat in the house daily in winter, the other set just to protect batteries.
Another point which I have shelved for the present is the use of a true diversion control for banks of solar panels. With diversion, any time your solar panels aren't pumping power into your batteries, that power is instead pumped into a heater or water heater or something that actually makes use of the power. This without pulling on the battery, leaving it able to concentrate on running the inverter to power the house.
And finally, my long range plans include possibly adding a wind turbine or 4 to the setup. I have a vertical 300w stuck up above the shop roof now, but so far receiving negligible output in decent wind. It came with one of those cute electonic controls with leds that blink encouragingly, but no output to speak of. Just took that out and wired in a regular 3 phase bridge rectifier instead. Now I'm waiting on a windy day.
I guess I should answer your question after all the above (More rambling). Why do I need dump loads? I don't really need them, but prefer to have at least one. When we 'Go to Town' we're usually gone most of the day. I want to be confident (as much as possible) that things are safe and under control while I'm not home. Redundant safeties make it easier to believe nothing serious will go wrong while we're gone. Bottom line, I feel better.
-
I heat water with PV diverting power from the panels DC buss with whatever isn't used to charge battery. My power point is about 60V and use the same 2,000W heating elements. That gives me up to about 500W. If you half the rated voltage, you get 1/4 of the rated power. Not a fan at all of heating water with batteries. Everyone thinks I will just divert when the battery is full. That is a very bad assumption. I sure can't tell when a battery is full. And you need a bigger inverter, bigger battery, and bigger charge controller wasting more money. Shame there aren't products to do this on the market. Dump load? I don't have any surplus of panels, just what I need to run the system. Still I find my water heater shutting off about 10am. I'm going to add another 30 gallon tank just for the excess. It is no wonder you are cooking elements, you can only dump so much into a tank. I just did some data collection over 11 days and it surprised me. Lowest temp was 107 over night. Tank shuts off at 127F.
I've done the oil finned heaters too. It is amazing how hot they get with just 300W. Unless the space is really cold, I would avoid putting more than that in them.
-
Hate to put it so bluntly, but if you can't tell when a battery is full, "you're doing it wrong". In a battery based system, that's in the top 3 list of things you best know how to determine unless you like throwing away good money after bad (or worse).
Dump loads by virtue of correct implementation do not "use battery" to DO anything. It's always surplus power. Just like when you "leave the lights on" in your car - with the engine running. The battery doesn't even know they're on; the alternator is doing all the work. You do not need a bigger battery to use a dump load. In fact, you're doing the same thing (just more directly) by diverting PV to the elements. Dump load, diversion load, shunt regulation... Synonyms.
Dump loading is just a "special" form of regulation, "usefully wasting" power that is otherwise available rather than letting the potential of which go completely to waste by not being tapped. A very useful and practical strategy when it's done right. Domestic hot water is an EXCELLENT place to put surplus electric power for multiple reasons. Like anything, yes, a water heater can reach capacity (and they do serving as dump loads) - If you're not using the heated water enough to naturally keep the heat down (which would appear to be the case the graph there indicates, most of the time displayed lol), the tank needs to still have a way to stay below a safe temp. So safety mechanisms need to be kept in place. Simplest method for this, aside from something like the standard switching thermostat, is to have the tank simply "blow off the steam" by discharging some of the water (usually somewhere protected from human or other access, outside, or down a drain that can handle it). Auxiliary tanks, yes, are another way.
Inverters are used to get the elements to operate at capacity, provide maximum dump ability, or both.
And yes, there are commercially available products to do just this... They're called... (Wait for it...)
Dump controllers. Diversion controllers. Shunt regulators.
Decent MPPT controllers either support a dedicated dump mode directly, have a feature to drive a set of contacts (mechanical or solid state, whatever floats your boat), or can even be "tricked" into running as one where neither of the other 2 methods apply.
I'm in agreement however on the oil heaters. That's playing with fire and hoping to not get burned. Better off with plain resistance in open air.
I've used wax personally with some success, but only with a purpose built supply to control it... The physics of the phase change can take a lot more energy input before overheating than the simple liquid oil in heaters.
No matter the route, a means of handling dangerous conditions needs to always be in play. Just like a battery, you can only put so much energy into any energy storage system before that cup overflows too.
Steve
-
Just about everything you said is wrong in some way.
-
LOL An amusing observation... In the mean time, the rest of us live in the real world. Thanks for playing.
-
Eidolon, to play devils advocate, im gleefully curious to ask where you think madscientist " was wrong " ??
-
And if you looking for reliable dump loads... try these, series them for voltage requirements ( bout 26v each to be safe) and they don't care ac/dc. Can be pwm'd etc. . The only downside is the very low initial resistance. .. but easily handle full 250w each...
https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.com.au%2Fulk%2Fitm%2F141410392736
-
Thanks for the comments and suggestions.
I'm not sure what your worry is about the oil filled heater. I have used one connected directly to my 48v bank without issues(thermostat shorted with a fuse). I read 5a with my dc clampon. Probably only doing 3a or less actual. The clampon is not too accurate. The heater warmed slowly and never got as hot as plugged into 120v (which makes sense).
[attach=1]
I have also used these direct connected to 3 series connected 320w panels in snow weather to help heat my shop and my camper. They get very warm at peak sun. Voltage about 95v drawing around 7 amps at peak sun on one element only. This puts the heater at about the wattage they are rated. Did this all last winter to see if it was worth messing further with.
In my well pump room I have one of these;
[attach=2]
It is direct connected to 3 series connected 190w panels. It has no thermostat so nothing to modify. It would get warm but not too hot at peak sun, and kept the regular heat lamp controlled by a freeze switch from having to work as hard. Keeps the well pump room warm without costing me as much battery bank power.
Most commercial dump load controllers I've seen can't do any significant wattage. That was my first problem. Most of the controllers that claim to be diversion controllers are actually just dump controllers. Another problem. Diversion is on the back burner for now. More important fish to fry.
Oh, and I like the little bulbs as dump loads. I will get some and try them out. More compact than the traditional big wire wound resistors. Easier to get and cheaper to buy too.
-
I also use an oil filled heater on my 24v system, ripped the original 2500w element out of it and put in a pair of 28v 1000w elements. No way can it overheat as it designed to dissipate 2500w and the max I can push into it is 2000w.
Works well (now I've fixed the leak!!)
-
I to have used these oil filled heaters.
As long as you don't cover them up or anything like that they should be fine.
The oil is just being used as a non corrosive means to spread heat though convection I believe.
It keeps the overall temperatures lower where anything touches them while still dissipating the energy.
Unlike the marketing says there is nothing more special about them.
A lot of solar concentrator heating systems use mineral oils as well, so I am not sure why the fears.
Am I missing something?
-
I'm now using 2 of the oil filled heaters in the basement, sometimes 1 heater only or both on with 1 element, sometimes all elements on on both. These are hooked to the 48v bank and I only run these during sunny days Measuring 5-6a at 55-56vdc for each heater with both elements on. No problems after a couple weeks of running all day.
Also installed one of the 400w ceramic panel heaters in the bathroom. It is also currently running off of the 48v bank. That one I've been leaving on 24/7. It only draws about 3 amps at 55-56vdc. Gets mildly warm and stays that way. All good until we get some no-sun days, 2 or more in a row and battery power becomes too scarce to waste.
Originally I had planned to connect that one to either 2 or 3 150w 95v solar panels. Maybe I should stick to that. Only warm in the day but no battery drain.
-
Hey guys!
Here’s my home made dump load for my windmill. Haven’t gotten that up yet. That’s next trip. Philippines is a long way off.
I believe Pete drew a mud map of this delta arrangement. I was thinking 3 phase would work best with a delta and he confirmed that and 1 ohm was the smallest I could get. They are 100 watt and with the heatsink compound and huge forced air cooled heatsink it should work. I have a 24v battery charger now so I guess I could test that theory. These were 5 for $4 so pretty cheap. 2nd set kicks in to halve the resistance if necessary. Heatsink came off an inverter control board for a mini split hvac unit. I used to get lots of these but they don’t send the heatsink anymore. Hell they don’t even give you new isolation feet for the replacement compressors either.
Maybe someone could line up some of these in a row on a heatsink like this from a scrap yard find until they get the right resistance and wattage and strap on some fans. I used nylon all thread and nylon nuts. Had more fans on it but those little fans don’t have a drop height if you know what I mean.
Good luck experimenting!
If it’s there, may as well use it, right?
-
Hello Posterity!
I haven't updated this in quite a while so guess its time.
Now that spring has sprung here in Idaho, we're getting a reasonable amount of sun about half the time. The hot water system works fine on moderately sunny spring days, and passably well even on mostly cloudy days. The regular hot water heater runs off of that 24v/8kw powerjack, powered by (4) 200ah batteries, which are charged by eight 320w panels. The pj puts about 110v ac at about 2500w into the original 240v water heater elements. Additionally, there is a timer that limits water heating to between noon and sundown. The water heater is 50 gal and well insulated so it holds temp well overnight. Of course if one uses all the hot water after sundown, that's all folks! (I can override the timer as necessary or switch the water heater over to grid as for example when we had company for a week) Normally with just my wife and myself, we have no issues.
Then there is the preheat tank. It is also a 50 gal water heater. Its hot water output feeds the cold water inlet to the main water heater. If there is sufficient sun to keep the 15kw powerjack running the house loads happy and the main battery bank is full, a dc controlled ac relay (dump load controller) powers up the elements of the preheat tank. It has 120v/2000w elements. A moderately sunny spring day easily heats up the preheat tank, sometimes before the main water heater is up to temp itself. When the preheat tank reaches a set temp, a circ pump kicks on circulating hot water through both water heaters and through the whole house piping, which allows the main heater's elements to shut off sooner.
So on days with any decent amount of sun we have at least 50 and usually more than 100 gal of hot water. If it will do that in early spring, then late spring, summer, and early fall should be no problem. Late fall and winter will have to use grid power most days. But thats no more than a quarter of the year so not bad. Even then we get some good sun days.
-
Here's a pic of my dump control for my pre-heat tank;
[attach=1]
Now that I look at the pic it looks pretty sloppy. A kludge. I need to dress things up.
The 2 white framed meters are actually voltmeters I'm using as power indicators. The upper one lights when the upper element is on and the lower is for the lower element. That way I know what's going on.
The breaker under that lets me shut off the ac feed to the tank elements. Black box on lower left converts 48vdc to 12vdc to power the dc/ac relay to its right. Bottom right is the actual 48vdc dump control.
Its all been working very well. (unlike my previous setup)
-
A month or so since last report. I don't think I switched to grid for water heating even once during April. Only twice that I remember in March. Early March. Now both water heaters are up to temp by late morning (unless someone uses a bunch of hot water during that time). It easily regenerates after one daytime shower. We're not even in full summer yet!
I'm pleased with the results so far.
-
Well, young man, that’s just Awesome!
Keep up the most excellent work!
Take care!
-
Been a long time. Almost forgot this place existed. This summer I ran my large LG washer off panels with no battery. Nice to just start a load, close the garage door and go shopping without having to watch the generator. I'm all about hot water. The house has two tanks powered by just excess solar not going to batteries. I had stored a 40 gallon water heater in the garage for a couple years and finally got that working to supply hot water to the washer only. That heater only gets excess energy not going to the other water heaters and battery. Hoped to get it up to 75 or 80F, but this often gets over 120F and is fed into the cold water hose of the washer. All cycles are hot. Chart is month long record of tank temp. This board keeps the panels at power point for heating water and can divert as little as 1W to a KW automatically sensing whatever excess is available. It can track power point and has arc interrupt so existing heater control can be used in current path. Indecently, the whole house works on a car battery and even have a dishwasher with heated dry.
-
Eidolon, I didn't see your post until I started to post myself. I have to say you are accomplishing some amazing results there. You are way beyond what I can do. Hats off.
My system seems pretty common compared to that, but its what I've got.
I just got a new 24v 15kw Upower inverter. This one is to replace my oldest and longest working inverter. It's still going strong but I had nothing else in 24v to back it up if (when) it finally fails. My experiences with trying to repair the older models prompted this purchase. Also, the older 24v is just an 8k model and won't power the water heater at 230v. Using it at 115v works fine but it takes much longer to heat up the tank.
With this new 15kw model I should be able to run it at 230v for a faster warmup. And my new larger Lifepo4 battery bank should handle it just fine.
[attach=1]
[attach=2]
So a stainless steel case is nice. More importantly, it has the new much simplified v10.3 control board.
[attach=3]
This control board can be jumper-selected for 48, 24, or 12v inverters. A big plus for keeping backup parts!
Regardless of what voltage is used/selected, the fan connection is only for 12v fans. This inverter, in keeping with powerjack's record of being inconsistent, has a 24v fan controlled independently from the onboard fan connection. Go figure.
-
As is my habit with powerjack inverters, after a quick visual inspection for shipping damage (none this time!) I run it for awhile on the bench. After that, its time to start modifying!
First, unbolted the transformer. Experience tells me the transformer will run too hot as shipped. So.
Cut some rubber spacers from a rubber buffing wheel (about $8 on ebay) and raised it up the 3/4" or so that gave me. This for better transformer heat dissipation. Later will mount a small fan on top blowing down into the middle.
[attach=1]
[attach=2]
Interesting to note that with this inverter, the transformer output voltage is back to 115/230v input is 18v instead of the 16v of my older models. Other previous powerjacks had 260v output voltages. I personally feel better with the 230v models as seems less likely to have damaging high voltage spikes.
[attach=3]
-
A few notables on this 24v 15kw inverter.
First, to be clear, there's no way this will ever do anything close to 15kw. This one is even chincier than the last one I bought. Of course I did not expect any better but thought it ought to be said.
Stainless steel case is a nice touch but unimportant to me. I would have been fine with painted black like before.
There is only one pair of battery terminals instead of two sets. All things considered, probably an improvement.
The L2 lead coming out of the transformer is very small. Appears to be just one strand of 12ga. I haven't yet peeled back the sleeve to see for certain. 230v loading may be a problem. The center tap is at least two strands.
Fuses in parallel. Whose bright idea was that? It doesn't "double the rating" if that's what they think. Actually makes the fuses useless for anything but a dead short of the load. I use external output breakers anyway, and usually just bypass the fuses.
A new mosfet used in this model I haven't seen before. NCE60H15a Rated to 150a. No idea whether good or bad.
The mosfets used in the 48v models are rated to 180a. Lower voltage for the same Kw means higher amps, right? I would assume a 24v model should have higher amp rated mosfets. Am I wrong? Between that and small wire sizes it is apparent powerjack doesn't expect this inverter to handle too much load for any length of time.
When I start pulling 4kw from a water heater I guess I'll find out.
[attach=1]
The last Upower inverter I got had solid heatsinks on the mainboard. No fins. It takes a strong fan to keep it cool. This unit does have fins so I expect better heat transfer there. Guess they learned something.
They are still doing screwy wiring, but not as bad as the last time. The L1 transformer output to the control board is connected to the neutral terminal. The neutral/center tap transformer output to the control board is connected to the L1 terminal and has the hall sensor on it (It used to be on the L1 lead). I changed it around on my previous unit but don't care enough to mess with it on this one. The output voltages measured at the output terminal are correct so it doesn't really make much difference, but why would they do it that way?
Last point of interest;
There is a ten switch pack on the control board.
[attach=2]
What are they for? Do they need to be changed if you change the voltage jumpers? This is something we really need some info on. We need a switch position chart! Oztules could probably figure them out in 5 minutes. Wish I was half as smart.
-
This morning added the small fan on top of the transformer and removed the battery charge input connections.
[attach=1]
While doing that noticed this very small board wired in with 5 leads and tucked into some heatshrink. Last minute revision? Its just two transistors and 4 resistors. Suspect something to do with the lcd screen.
[attach=2]
[attach=3]
Added a couple refinements to my preheat water heater controls.
[attach=4]
Bottom left white 24hr timer lets me designate window of operation (daylight hours). Above that is the voltage control to limit operation to proper battery voltage range. And added small fan to keep solid state relay cool.
The two whjte voltmeters at the top indicate which element is on (mainly so I know what's going on at a glance)
With this stuff done its all pretty much automatic and ready for summer 24hr use.
-
Well, I put the fan on the transformer, then the cover wouldn't fit because of the lcd screen. Ok, the fan is more important than the screen, so I reversed the hardware and mounted the lcd on the outside of the case. A little hokey but don't care much about the lcd anyway. This unit is intended for my water heater setup, so I hardly ever look at the lcd anyway.
[attach=1]
Oh, well
-
As I was setting this inverter up on its shelf, thinking I was done modifying, I remembered something. These Upower units are all using 110v control boards. which is fine as far as it goes. But.
If you are going to use the 220v or use the L2 leg for 110v you need to add shaping capacitors to that side. Doesn't really matter for my water heater use, but if you look at the sine wave of the L2 leg to neutral/center tap, its very fuzzy.
Since I want to be able to use this inverter as a backup to run my house off of the 24v water heater batteries, I need to add a couple in.
A look at the control board shows the 2 large brown caps across the L1 and neutral lines.
[attach=1]
[attach=2]
Powerjack used to use 475j 630v caps, 2 in parallel on their 110v boards. All the newer 110v boards use 475j 400v caps (physically smaller). On the older 230v control boards, they used one cap across L1 and L2, which seemed to work fine.
I have some of the 630v on hand so that's what I'll use.
[attach=3]
Since the L2 lead from the transformer goes directly to the output (no connection to the control board), I'll just solder a short piece of wire to the two caps in parallel and connect the wires to thr L2 and N output connector. (I thought about removing the two caps from the board and using one across L1 and L2, but that's too much work.)
-
The new inverter for the 24v bank is on the shelf next to the old one and ready to go, but...
Being lazy, and reluctant to change what's working well, I'm going to continue with the old inverter for now anyway.
If it fails I can hook up the new one in a few minutes.
I have used no grid power for water heating in more than a month. With longer, sunnier days coming don't expect to use grid power unless we get bad weather. Maybe we can make it all the way to next winter.
-
As we move into spring and summer, the 24v water heater bank is working so well I almost forget to check it. With 6 of the byd modules, if they were new, I'd be calling it a 1200ah bank. As is, maybe half that in reality. Still way more than needed under normal conditions. It's those wintertime 'normal conditions' that are at issue. I'm hoping to add maybe 8 more solar panels to this bank's charging, all at high mounting angle for winter. Problem is where to put them? About every viable spot is already in use. That leaves my crappy garage roof. I have to face removing the crappy existing sheet metal roof, sheet it with waferboard, then roof it properly. Then I can mount some panels on it.
With my already lengthy project list, don't know if I can do it this year or not....
-
Hi Doc well it sounds like you have it sorted.
Just a short note on waferboard. From what I have seen it does not stand up too well to condensation. Our carport roof was lined under the steel sheeting with MDF board, it has swelled and sagged quite a bit from moisture getting into it.
If your garage is closed in then it may be fine to use waferboard but watch the humidity and moisture.
Pete
-
Thanks for the heads up, Pete. I'm certainly no expert. Just thought that's what everyone used. My 'garage' is a converted carport, so not very tight and that's why it has a crappy sheetmetal roof (that leaks in heavy rain).
Any better roof will be an improvement, but don't want to waste my time and money putting in something that's going to fail soon. What do you suggest?
-
Hi Doc the most weatherproof lining for a carport or shed roof would be Cement Sheet. You could get away with tongue and groove timber boards but make sure you have a layer of foil or some other sarking over the boards. I think that the mistake the people who built the carport at my place was that they did not use foil over the top of the MDF. Then again Foil or Sarking can cause problems later down the track if rodents eat it, which they do.
So if I ever take the roof of my carport and fix it, I will be using cement sheeting with plenty of supporting battens so that it doesn't sag. Sometimes if it is just laid over the rafters and the gaps are too wide it sags a bit too so thin timber battens help prevent that.
Cheers
PEte
-
3 days of dark clouds and rain. Had to switch the 48v main bank back to grid overnight. No problem with the water heating setup though.
The 24v water heater bank sailed right through 3 cloudy days/nights without much trouble. It is performing very well. I'm happy with the decision to make it a (mostly) separate system.
-
Been awhile since I've had something to post here. Water heating has been working well for months. Now that winter is here and sun is more scarce, added 4 winter-only 325w panels to my 24v bank (for water heating). Helping a lot.
As the final winter-only addition, I just finished adding (4) 265w panels that are powering the lower element on my preheat tank. Put in a transfer switch so I can go back to normal when I want. Connected the panels to this DC-DC boost board, then to the element.
[attach=1]
The idea is that with this both the top and bottom elements can heat at the same time.
Only had it in for 2 days so far so the jury is still out. Not sure the boost is doing any good. Might play with settings some more. A guy on DIY Solar used it and said it worked well for him. It's a 50gal tank so maybe I'm just expecting too much.
Will give it more time....
-
I'm still around. Been trying to log in forever since my old computer died and couldn't remember my password. "Forgot my password" routine hasn't worked even though it recognizes my email. Finally found a scrap of paper with old password on it.
Im still heating water and this year I started desulfating old batteries I get at town recycling with the water heater as it conveniently produces 10A pulses. The garage system got two more panels added to it which face east. Greatest move ever. I can now run the LG washer early in the morning with no batteries. [attach=1] Note the jumper cables.
So what boost converter is that? and what are you boosting to?
-
Hi Eidolon,
Good to hear from you. The boost controller is a
BST-900W DC-DC CNC Boost Converter 8-60V Step-up 10-120V Solar Charging CVCC
Here's a link for it if you're interested;
https://www.ebay.com/itm/224529024087?hash=item3446f90457:g:T7IAAOSwrolg6ZoK (https://www.ebay.com/itm/224529024087?hash=item3446f90457:g:T7IAAOSwrolg6ZoK)
There are no directions for setting it.(typical chinese doodad!) Figured out I had to set the constant current to zero to get it to do anything. I've only got constant voltage set for 60v at present. Been pretty cloudy, but when there was some sun it was doing between 8 and 9 amps connected to the (4) 265w panels. (two in series and those 2 sets in parallel) Soon as I get a sunny day I plan to bump it up to 100v and see what I get there.
Probably forgot to mention the tank elements are 120v/2000w elements.
No progress today as we were fogged in all day.
-
With all the gloomy weather, I can't really say if the DC-DC boost is really worth the expense/effort. It puts up to 9a into the lower heating element so must be getting something out of it, but it's not much. The best part is that it is independent of everything else, so works even when grid is powering everything else.
The real question is whether just connecting the solar panel output directly to the element would get me more water heating than using the boost controller. I'll give it a few more days like this, then try the direct connect.
-
Hi Doc! Looks like your doing ok, seems better than i wouldve guessed. So you have two 30V panels in series= 60v + a little for the cold... for a current of 8A x2 = total of 60v times 16A = 960W available in full sun. You are boosting into a 125V element at 9 amps. Without knowing the voltage at the element, 16A x 125v. = 2000w = 7.8 ohm element. E = IxR ... E = 9A x 7.8ohm = 70.2V. P=IV so P = 9Ax70.2= 631W delivered with the booster. Direct connect the way they are will be P = I x V = (60/7.8)x(60) = 7.7A x 60V = 461W, now if you change the element to a 1000W 15.6ohm element and connect all 4 in series, then you will deliver 960W+ cold bonus probably a bit over 1kw in full sun, you would then loose some in clouds when panels cant do imp. It why i quit doing it when my grid tie capacity was capable as they automatically do this conversion you want. Ive not played with the boost or buck to do this. Both circuits can do the conversion but im not sure they can in stock form as they are designed to control the out not the in. The boost converter current limit perhaps offers a way to cheat to keep the panel volts up during clouds but you need a way to turn it up automatically when sunny.
The simplest most troublefree way is to match the 125v 1k element with the 4 in series.
It looks like your only loosing 300-350w in full sun, if your panel volts are staying above 25 during cloudy weather, you are acheving some kind of crude mppt win. If not then matching the 1kw element with the series 4 will be superior.
. I like wood heat and a plate heat xchgr in the winter. Those babys can make hot water as fast as you use it.
All the best!
-
I played around with a boost converter to raise 60V to 80V using a China module modified to keep the panels at near power point. This was just an experiment to match the element better before I installed the new one. The booster really cooked, efficiency dropped under 80% and the whole effort only gained a few extra watts.
The main problem with buck or boost converters is they easily throw the panels into a death spiral. To meet their output, current from the panels has to increase and that drops voltage even lower. I have a circuit that can be added to any module that will keep the panels from dropping below a minimum voltage if interested.
-
I have a circuit that can be added to any module that will keep the panels from dropping below a minimum voltage if interested.
Definitely interested! Would this be instead of or added to the boost module?
-
Hi LH,
I have been thinking about what to try next. Was next going to just bypass the boost module and see what difference it makes. Then thought I would re-arrange the panels in series as you suggested and give that a go. I know I'm getting something from that setup, it just doesn't seem like much. Wish I could add more panels but no place to put any more for this application. (Can't just stick panels out in the yard with all these high winds we've been getting! -- They have to be tied down to something solid.)
-
Hey Doc!
Yeah your not kidding there about the wind, we had 100mph wednesday evening.
Ive no idea how i didnt have damage. Tore up some roofing and skylights where i work. Nothing that wasnt repairable but it was not just a breeze. I think the trees blocked most of it here though.
For several years i had 7 x 72 cell panels in series running a 1500w 240v element 38ohm. (Ran at about 1800w). It worked so well except two things, cloudy days and it killed heating elements sometimes 6 months only. So i got a 2000w 277v which is still 38ohms with lower power density. Never burned it out.
You are almost there with what you want. I could see you leaving your 60v parallel connection to the 125v element and adjusting your current limit for full sun vs clouds either with a switch and resistors or a relay. If your around home flipping a switch for clouds is a simple solution.
-
I would like to start out saying that I think boost converters are generally a bad idea. I had someone from Europe inquiring with an array of 120V. Their water heater had a 2,000W 230V element and there was no lower resistance replacement for it. In that case using a 48 power supply to increase the voltage was justified. It was not an issue of efficiency, but accomplishing a goal.
A boost converter can not be totally turned off. Input voltage always flows from the input to the output thru the diode and inductor. Only the amount of boost is controlled. The lower panel power is just passed thru. One has to accept that the lower wattages are just lost. I give this example which can control a boost or buck converter only for educational purposes.
Every switching converter has an IC pin which is a reference which is connected to the output to regulate voltage. The reference is typically 1.2 or 2.5 volts. That connection is easiest to find by either measuring the lowest voltage or resistance. Pot pins are easiest to solder to. When the reference pin goes above the reference voltage the switching regulator will turn off. A 12K resistor can usually supply enough current thru the diode to that pin. Less than 1mv increase is more than enough. The diode is there only to affect the circuit in turn off mode. When the FET, TL431, or opamp pulls the 12K resistor to - common, the added control circuit has no effect,[attach=1] and the switching regulator operates normally. A FET is used in this example for simplicity, and it is a power FET because these are common. It is also not linear having a sharp turn on between 2.5 and 4V with a high impedance that will not load down a voltage divider. There is some drift with temperature. Everything to the right of the diode are switching module components. The 12V can be obtained from the regulator of the switching module.
As PV voltage drops below the power point voltage, added current feeds to the reference pin. This adds to the current from the output voltage divider. This action does not completely turn off the switching controller. Rather it reduces the amount of boost till the PV voltage stabilizes near power point.
This circuit helps a little preventing a boost converter from putting the panels in a death spiral by drawing more and more panel current. It turns off the boost converter and makes it the same as direct connect. Boost converters are just a bad bad bad idea. To be efficient the panel should feed into a capacitor bank and be pulsed from that maintain a constant voltage from the panel near power point.
-
Here is a version of one of my designs in northern Europe. I'm quite impressed as he had no prior electronics experience and even etched his own circuit board. He states he is getting over 5KWH a day in winter from four panels. Please respect his privacy and do not comment.
Lithuania video
This next pre shipment video is of a board I sent to NM operating on four 255W panels facing east. This is on the 40 gallon tank in the garage I just use for laundry hot water. It is
connected to only the lower 5500W element for the 120V test. I normally connect upper and lower in parallel with my 60V house array. Setting priority on the basis of PV voltage works well at only diverting excess house power to the garage only when house water heater finally turns off. The garage has a transfer switch so I can run the washer off 120V DC garage panels without using a battery. Totally free laundry now. The generator is only used now for the wood saws.
120V first test
-
Here is an example of constant voltage as represented by the green line. The small higher peaks are when the panel goes over power point due to heating element being higher resistance than optimal. Power is the yellow line.
[attach=1]
This other graph is when the board is operated in diversion mode. This scan is only a half hour and demonstrates how varied power can be. Besides clouds, as devices in the house turns on diversion drops to zero. At the very end the voltage drops to 44V, likely refrigerator turning on. Manel panels are shaded reflecting the lower voltage of the charge controller.
[attach=2]
-
You guys take a pretty high tech approach, but it seems to be working out fine.
I have a evacuated tube solar hot water system. It is backed up in winter with a wet back on our wood heater.
Then when the wood heater is not running and the sun not quite strong enough, I use an element in the tank and power it by my inverter.
It is a low tech approach where I look at a guage I have to decide if the water temp is high enough. Then I switch the element on. The power is regulated by both a thermal switch on the gauge and also by the battery level.
The Victron Battery monitor I have has a settable relay, so i have the relay switching a higher power relay to turn the water heater element on. It allows the element to run only when the batteries are fully charged and turns the element off if the batteries go below 97%.
I know it is a low tech approach but it suits me.
I am guessing that your systems boost DC up and use DC to run the elements, Does this damage the over temp thermostats in the tanks?
Cheers
Pete
-
Many focus on a single issue. Heating water can be complex having many issues. My board has arc interrupt so that high voltage DC can be used with standard mechanical thermostats. Water heating should be done with the highest PV voltage directly from the panels eliminating conversion losses and avoiding buying more expensive equipment for the increased load. My method allows using existing array to heat water and supply charging needs which have priority. These methods will be common in another 10-15 years. At this time almost no one understands the science. For now the solar world goes for easy.
-
My board has arc interrupt so that high voltage DC can be used with standard mechanical thermostats. Water heating should be done with the highest PV voltage directly from the panels eliminating conversion losses and avoiding buying more expensive equipment for the increased load.
I admit I don't completely understand the specifics of your system, but agree that eliminating the converter should be more efficient. Also the idea that your board makes using panel voltage safe for water heater thermostats is a big plus.
Do you build your boards for sale? I would like to use my four panels in series for an open circuit voltage of about 120vdc. Would that be compatible?
-
agree that eliminating the converter should be more efficient.
The exception to this is where you are using an element that is not well matched.
Lets take a really BAD example...
You have a 5kW, 11.5 ohm resistive element (ie, 5kw/240V element), and 2kW of PV panels connected to give 60V output.
In full bright sun, you would think "I should be getting 2kW into my tank"... but at 60V, you're only going to convince 5.3A to flow for a total heating benefit of just 312 watts.
Using a dc-dc converter, or dc-ac inverter and boosting that to lets say 120V, even if your inverter is only 90% efficient, you're going to get 1125 watts into your tank.
1125 watts is almost 4 times more USEFUL POWER, even if you are "throwing away 10%".
You need to look at the big picture before deciding what is the "best" way to do most jobs.
-
Well, I do agree that most everyone just slaps a system together with whatever they have and just hopes for the best.
What I want to get across is that every successful solar system has excess power by design in order to recover from bad days. While many concentrate on getting that extra 2% efficiency by buying a better charge controller, they throw away thousands of WH not harvested.
Many with camp systems live like refugees when they could have ready hot water with minimal added expense.
Whatever system is used, monitoring PV voltage is an excellent way to divert excess power to water heating.
-
The easiest way to explain the board is to say it is designed to use four grid tie panels. These will be in the 7-10A range and the heating element which will match that, 4500-5500W 240V heating element. This is what similar products are designed for. The capacitor bank is designed for a certain RC time constant. The capacitor bank must be able to supply nearly full power for 1ms. My camp system operates at 60V which is the extreme low voltage end region of operation if the array can supply 500W most of the day. I only talk about it because it is possible to get acceptable results with 120V heating elements in camp systems. Many tanks do not have replaceable elements. My board is designed to operate in parallel with charge controllers harvesting excess power. To my surprise there is no interest in this. Everyone wants stand alone systems. This board is designed to use the natural power point of an array
and will not work well with single string partially shaded arrays. I would love to have a 90 or 120V system, but shading prevents that.
I had requests to create a board with high reliability and a high tolerance from lightning spikes. Techluck boards were failing and that company would not repair them. Proprietary microprocessors were avoided in favor of common discrete socketed IC's which would be
available for the next 20 years. The FET can also be replaced with a screwdriver. The capacitor bank is far more robust than other products for long life. Techluck uses only three 100uf capacitors which are rated at .58 amps. By contrast I have five 1200uF capacitors
rated at over 3A each and the design operates at lower frequency for less noise. This allows others to copy the design with simple perf board construction and still have high noise immunity.
A limited number of the first run evaluation board and is available for purchase. It is unlikely there will ever be more made. It can be as low as $70 + shipping. Initial board only cost is $140 and $70 is refunded after you post an evaluation with a video. Sorry I have to do this, but not one person has contacted me once they purchased a board with test results. I use these boards at two locations that have extensive shading and can not compare results to expected results of a solar calculator. These boards were never intended to be sold to the general public. Rather, thru installers who would take the time to learn how to set them up properly. You must also supply a photo of a power meter before I ship. Module can not be set up without this. Array / heater element information is also needed to insure proper matching. I assume this is just too much involvement for most people and I recommend for them the ACTii from Poland for about $200.
-
Hi all! I had two thoughts on the water heating thing. Doc mentioned a preheat tank. If that doesnt work for you, another approach that i use is a 100 gal tall electric. Not everyone has thought about this but water stratifies in an insane way. The water at the top can be 60C when the bottom is 20C and it seems to stay that way. By running top and bottom elements on different thermostats, you can use a priority heating scheme to meet your need. Also, i have a well shed with 80Gal pressure tank. Ive a 250w heat lamp plugged into a 3°C switch sitting beside a vulnerable pipe. So ive kinda the opposite scheme as doc where this water starts out at probably 10C drops to 3C and then gets piped to the house then heated in water heater. So wouldnt it make sense to move that pressure tank to our basement where its around 25-27C? Or would i create more problems like a pressure switch that cycles quickly because of delayed pressure response. Cant say ive ever seen anyone set up this way but then not sure anyone would have a reason to.
Hope everyone is getting plenty of sunshine! We got about 27kw today, yesterday was good too!
-
Ah, water heating! I've spent a lot of time and money trying to get the most I can from my system. Mostly it's working very well. This year for the winter months, I'm putting the four panels of the DC-DC converter all in series instead of the series-parallel setup used previously. I got a moderate amount of water heating from it last year, but I think it will do better with all 4 in series. Double the voltage input to the converter should increase the amount of time each day when the converter sees usable voltage to convert.
Sounds good anyway.
-
I will have to bypass the converter and hook the panels directly to the element of the water heater with them in series. Series connection is too high of voltage for the converter. Minor detail.
-
Wouldn't an MPPT accomplish what your trying for here, or am I missing something?