The panels can power the setup during the day the battery is for times when the panels dont produce power. So you need to have enough panels to charge the battery and power the setup.
I have a single 100W panel. For three days in a row now, it has produced <10Wh. Of course this if Finland, but I wouldn’t trust a solar panel for devices that absolutely need to be powered on 24/7 without a huge battery.
Pssst! Maukka see my signature (your image doesn't meets mobile friendly standards). :-D
Thanks, that was enlightening. Looks like the above panel only worked at a good pace for a single day out of 7, with another two honorable mentions.
Spoke a bit more with my colleague. Apparently the pump setup is to move water between adjacent ponds or bogs, and doesn't need to be powered 24/7. This is quite a difference.
The above graphic is pretty useful, lets me know the panels will only be able to power up the pump setup for a few hours a day, if anything, without a battery. I can now see the problem much better, namely because it is not coral, fish and squid life critical. 
Sat, 10/19/2019 - 11:11
Get the Max Watts to Dollar you can find, S/H 250-300w grid tie panels are ideal giving 3 times the power of most 100w panels for the same price, what ever you think you need triple it
cheap PWM controllers can cut panel watts by up to 50%! (look up MPPT) its a tuned buck conversion far far more efficient
Lead acid is ok if you get GEL batteries , lion is better but massively more expensive.
On a good day my little hobby system produces around 1kw for 5-6 hours generating roughly 6Kw/h it stores 2.5Kw/h of that for night time and sends the surplus to the Grid tie… on a bad day it can make as little as 0.2kw/h for the entire day, during the winter i don’t even bother to cycle the batteries i just fully charge them and switch panels directly to a grid tie.
230AH 24v Gel Bank / only cycled to 50% dod giving 2000+ cycles.
Here’s what a winter week can look like!
Wind turbines can help, there are some quite reliable kits coming out of china that can hit around 400w Turbine Video review
The_ Fat_Controller That is a beautiful solar setup. Done neatly and efficiently. Good work!!!
Hello again. :-)
My colleague told me the “potential customers” aim to lay some fresh water fishes in the ponds (!), and they will see how much will it go without water circulation. The stuff will have maintenance every two days at least. I am no aquarium expert, so don't know how senseless or senseful this is.
Yes very nice setup The_Fat_Controller, that should allow you to save quite some money in electricity bills. Thanks for pinpointing the above concerning MPPT controllers (so much abbreviations which God knows :facepalm: what they mean without proper context).
The above panels are good, or so it looks to me (the bulk of this message was writen before hank's latest contribution). The question here is more like how many of them will be required, either two or three for the time being.
In the meantime I looked for some no-frills, affordable MPPT controllers:
- https://www.aliexpress.com/item/33036101920.html Sunyima MPPT boost controller for up to 300W of panels (from AiyimaElectrical, some feedback; you may find it somewhat cheaper at the next seller).
- https://www.aliexpress.com/item/4000170651351.html Sunyima MPPT boost charge controller for up to 600W of panels (from Sunyima Energy-Saving).
I aim to DIY the battery pack out of either 18650 li-ion or 32700 LiFePO4 cells, at least for the time being (I may change my mind). 
Things to be aware of with cheap controllers , Many sellers lie about them being MPPT most are in fact PWM there are many fake Mppt modules
there are also some good cheap controllers which really are MPPT
here is one example
Real budget MPPT
You can tell because internally it has Inductor
the other give away is the conversion if you feed mppt say panel voltage 24 @ 1amp… you will Get 2amps at 12v , with PWM you will just get 1amp @ 12v
Here is another good cheap TRUE mppt
Boost Mppt
The issue you will face with most of these is not conversion efficiency but very restricted charging options, many do not have voltage sensing and reliy on preset values with timers ok for floaded lead acid but not much else
Others such as the Epever tracer Tracer Series offer more control and can be setup for custom charging routines I believe these can be setup to safely and correctly charge lion cells they also support load shedding and remote access, the epever is a good controller and i used one for quite a while (there maybe better options available now)
The Victron controllers are probably the best they have full customisation of the charge routine and can be programmed for any type of cells such as Tubular Gel li on, lifepo4 etc , bulk absorption float etc and it is all monitored on battery voltage with full customisation
Victron is surely the way to go if you just want it to work and make configuring easy. It’s very handy to monitor locally or over the internet with a Raspberry Pi running Victron’s Venus OS.
Mine is just in the testing phase, since I don’t have a good location for the panels.
Bottom line with panels is Watts are watts and the more you can get per dollar spent the better , being more efficient simply means a smaller panel gives the same watts. but if it costs you 3 times as much per watt its pointless. Large industrial panels are built to last many years mine tested to 275w each (300w rated) and cost about 100 bucks each 
The ones you point out on alliexpress are not bad if they really give 200w for 120 bucks but id check the reviews on them not all panels are created equally and price doesnt always mean a lot
NOTE: you will almost never see the full rated output from panels unless they are below 25C. people always assume you get max power in really bright hot sun but this is not the case. as semiconductor temperature rise efficiency drops (solar companies neglect to mention this)
I love the victron Its very good controller but OWch! the Price sucks!!
PS: for good cheap inverter option PURE SINE , look into APC Smart UPS can be controlled vier Virtual load output on controller. (Epever also) I used this method initially but found grid tie inverter to be more suitable for my needs but if i was going offgrid I would go straight back to the APC
There is a guy on Youtube by the name AverageJoe , he has many detailed videos on making and protecting Lion packs as well as balancing & monitoring well worth a watch
I don't have any direct experience with this pump. The PV version of it can be connected directly to solar panels without separate controllers or batteries.
https://www.thesolarbiz.com/el-sid-sid10pv-10w-pv-direct-pump.html
Sort of nailed it, :-) thanks. For just a tiny bit more money than the aforementioned Sunyima it offers better interface and looks like a better overall machine, even though it is fan cooled… 
And I really mean just a tiny bit more expensive: https://www.aliexpress.com/item/32931442381.html
The panels are a bit more expensive , I see €142.16 for a pair and €209.49 for a 3-pack @ solarfactoryretail.
What the… wow! That pump is @#$% expensive! 
Thanks to everyone, for now I have plenty of data to start with. This project will load the battery with buck or buck-boost converters and DC pumps but pretty sure I'll take a look at the inverter thing sooner or later. Many of the “high end” device suggestions will come into play once I am more familiarized with the stuff, I presume.
No worries , those little 7210a’s really are excellent for the price only issue is they are boost so battery bank needs to be higher voltage than panel but its ideal if you want to run a 24v bank with an 18v pannel
here’s a review
Wellp, while I am more or less fluent in read and written english language, the same I cannot say concerning spoken english understanding. I may have certain negative beliefs in this regard, forged at an early age.
Despite this, looks easy to understand. In red is the input voltage from the source (panels, wind turbines, mini-stellarator or mini-tokamak, etc.). In purple, with yellow V is the settable battery output voltage, and with yellow A the settable maximum output current. Below and all in yellow is output power in watts. In cyan colour figures battery capacity can be specified (what the point is right now I don't know), and a timer count whose point is also sort of pointless :-D to me. At the right side column I see a diskette icon with a two digit counter, I presume that is a writeable¹ array of memory presets…
I also have a small concerning doubt with the MPT-7210A, will it be able to support 16.6̅7 amps of input? I know I should stay below 10A at the output (for such I thing I would choose no less than a 9S li-ion battery for 200W of “18V” panels, and a minimum of 13S for 300W), but no figure I've found yet concerning input limit. The above Sunyima boost controller specifies 16.7A maximum input current, for example.
¹ Would you say writeable or writable? 
Generally the max current is regardless of voltage , as such most controllers will handle more watts at higher voltage
For example my victron is rated for 30amps , if i run a 12v battery bank my PV max will be 440Watts , but if i run in 24v @ 30 amps i can have upto 890w of PV and so on
the MPT-7210A is 10amps max input or output ,so you’d need to go 48v (13S) bank and run your solar panels in series to give 36v that way you never go above 10amp on the charge controller.
The issue then is getting from 48v back down to the voltage you need for the pump so you’ll need a converter
And vice versa. If you live in an area where winter sunny day temperatures can drop well below freezing, be aware.
An old friend living off-grid once boiled his lead batteries dry in such cold conditions because his solar panels got so highly efficient.
Well, for now I can say this thread gained a lot of momentum. It's been a busy O:) weekend.
Guess that arithmetic is maximum charge voltage times maximum current, yet these are maximums and you don't want to be pressing the pedal to the metal frequently, or do you? Long term reliability matters.
Well, I discern the MPT-7210A is then not really well suited to handle 2 × 100W “18V” panels in parallel, as they can give out in excess of 5.5A each.
Given this and considering I may go with up to 4 × 100W panels, the better affordable solution as I see it is to choose 2x (Sunyima) CTK-EV-300 charge controllers, each one fed by 2 × 100W panels.
By the way, looking into this I found the following page with specifications for quite a lot of the devices discussed here: Product list @ Nanjing Longline Electronic Technical Co., Ltd.
I really like the DPS3806 boost-buck converter, affordable and flexible device. Found a video of it made by the same Julian Ilett fellow which talked about the MPT-7210A above:
SEPIC topology maybe? They're easily parallelizable. 
[quote=Barkuti]
Well… in reality it is very very rare that you will ever see the maximum from a solar panel unless as someone else said it is Freezing cold and you have very very strong sun , the ratings they give for Pmax assume absolute perfect conditions, Most sensible controllers are also current limited so they will simple draw up to their max rating rather than self destruct
[quote=Barkuti]
correct it would be 2 panels in Series… Parallel would be to much current also no head room for later upgrade , it is always better to get a larger controller if you can
There are some very nice Boost buck up & down converters around you dont need to spend alot of money to get the power out of the panels, the issue is as i said earlier going to be safely charging lithium batteries
Im not sure of the requirements for charging Lithium cells … for my Tubular Gel batteries they require a current limited 3 stage charge , Bulk Absorption & float and each voltage has to be temperature compensated (if you cause a gel battery to GAS) it is ruined and capacity lost so they must never be equalised
Im no expert on lithium but id imagine at the very least they will need Balancing, current limiting & over voltage protection , I think i also read somewhere they must not be charged if temperature below 0
I would have preferred lithium but cost was an issue and price difference big 80ah at 12v Lifepo4 , VS 110ah at 12v Tubular GEL …
Lead acid batteries cannot offer anywhere close to their 20 hour ratings if discharged fast, much less the above 100 hour rating albeit this may be inconsequential due to battery size versus average power consumption. I also wonder about their actual lifecycle performance, but looks like they're still the most affordable technology.
I've already built a few small li-ion batteries with modern 18650 cells (LCO and NMC chemistries), going as far as binning and grouping cells to maximize series pack balance and lowering maximum charge voltage to boost cycle life too. No issues with them, guess it is too early to say anyway.
If you are willing to take a chance with some cells from china the price for li-ion can be lower, although it requires battery pack assembly skills. Take a look at my top choices at this moment:
https://www.aliexpress.com/item/4000197218968.html
This is a cell from a still unknown to me OEM, and rewrapped by some. Here's a review of it: Queen Battery QB26800 6800mAh 20A - the big guy! @ Thunderheart Reviews
By charging that cell to slightly above 3.9V still above 20Wh per can can be extracted at a low C-rate, which is pretty interesting.
https://www.aliexpress.com/item/32795814865.html
As far as I understand this other cell is from Heter, it is used by some Ukrainian fellows in their DIY e-bikes and unorthodox tests. Tester Henrik will soon publish a review (lygte-info.dk).
A problem I face with cell balancing is that commercial balance boards are designed to do it at very high voltage, above maximum permitted voltage. This is not ideal, in my honest opinion. It probably is better to install switchable voltmeters at each battery stage together with pushbutton enabled bleeding resistors (inexpensive flashlights with linear drivers would probably be a nice idea :-D for such role), and do balancing by hand from time to time.
By the way I've noticed that DC solar cable can be somewhat pricey. I wonder if some beefy silicone wire could also do good… 