diy solar

diy solar

Water Production Plant

pietropavan

New Member
Joined
Apr 5, 2020
Messages
2
Dear All,

I am running a potable water production plant, composed of several stages of filtration including a double reverse osmosis system.
The whole system is running on 3 pumps with a total estimated consumption of 4KW.

The pumps installed are now AC single phase 220V and it is grid connected.
pump1 = 2HP 1100W
pump2 = 1HP 750W
pump3 = 1HP 750W

I would like to try to transition to solar but have 1 gazillion questions on how to make it work.

First thing to understand would be: Do i need to change the pumps to DC and run without inverter ?
It would be obviously preferable to leave the existing setup unchanged and run with an inverter but will it be able to cope with the spike current of start up ?

I know most inverters can handle spike startup currents but any miscalculation looks like it can be fatal...

Any ideas how i can design such a system by using inexpensive hardware ? LiFePO2 batteries vs Litium... BMS ?

Thanks in advance for your help.
 
You can use an inverter, but I'd recommend putting soft starters in front of your pumps to limit inrush current spikes.

A 4kWh consumption would mean that, in order to run this independently for one hour, you'd need a slightly larger than 4kWh battery if we stick to LiFePO4 (this is a lithium chemistry by the way). This is very much doable, check some of the threads on this forum discussing 280Ah cells form Xuba/etc. through Alibaba to get an idea on cost. Aim for a 48V battery though. Yes, you would want a BMS - there are several threads discussing various options; I suggest you check those out first.
 
Last edited:
Using my own system for water pumping, I have a set of questions for you. First, I'd like to ask by what methodology are you coming up with your wattage values? Looking at your values, I question the accuracy of the numbers. I myself have a 1hp 240VAC Grunfos pump, and that's pulling about 9.5amps at 240V, which is ~2300W. These numbers are coming from a Fluke clamp meter with "inrush current" capability. I'm finding that the Grunfos pulls about 38amps right at startup, or 4X the running amperage. To me a 2hp pump drawing only 1100 watts is suspect?

Secondly, will all three pumps be running at the same time, or do they function in stages? Are they likely to all draw their surge currents at the same time, or are they staggered?

Third, what kind of electronic monitoring will be involved in water production? What requires AC besides just the pumps running. Will sensors monitoring water quality also need to be supplied with AC?

If going the AC route, I'd look at Outback's Radian 8000, and Schneider's 6848, which will both produce AC in the range you are looking at. I would expect a 48V battery bank in at least the 400-500AH range will be needed.
 
If the pumps startup can be staggered you can get away with less inverter.
 
Hello, thanks for the many replies.

First of all i have to correct myself on the 2HP pump. This is a mistake. There is nothing written on the HP on the product placard. This is a 13 Multistage pump made in China, Model CDLF1-13, P rated at 1100W as written on the placard. i cannot still figure out the HP as the manual is super confusing.
In the comparison with the Groundfos pump, that is a much better product, i believe its made in EU, and you are rated at 9.5A while my chinese pump is rated at 7A only at 220V. So if the math is not wrong the power absorbed should be 1540W and not 1100W as its written on the pump placard... no idea why they write 1100W on the placard !!!

I am currently not at my business and trying to get the number out of the confusing datasheets of the products i have here with me. The 4000W was just an estimation. I am asking my staff to send me some pictures of the actual pump plates so i can be more precise...

I understand the safest thing to do is to measure all spike currents at startup as suggested... I would like to ask more details on how to measure the startup spike current, is an ammeter just enough or do you need something more like a logger or some other instrument ? I would have to take these measurements as soon as i get back to my business, but in these times this is difficult to tell...

In relation to the operation of the system, unfortunately it is designed really bad (not by myself), and yes the 3 pumps are allowed to run at the same time. I am actually studying a way to monitor and control the whole station through PLC logic. I was looking at Beckhoff PLCs as well as AutomationDirect BRX series, but i found that company extremely difficult to deal with.
I can create a logic where i can limit and stagger the operation of the pumps but this has to be studied appropriately and the first step is to acquire data on how the system is running... So i need to rig the station properly with flowmeters and pressure sensors first, in order to be able to come up with a logic, based on the volumes of production we have.
At the moment there is no kind of any electronic monitoring but if PLC are added that could be another 40W to 100W requirement depending on how many sensors i put... but for sure nothing as big as the power required to run the water pumps.

Also interesting as i did not know about the soft starters, i read the link provided but if anyone could suggest any names of companies or something to get started to research that will be appreciated...

One more thing, can i connect inverters together in order to double up the power ? If so where can i get more info on this ? I have seen that some manufacturers offer a possibility to interconnect with serial or a special comm bus in order for the inverters to work together and optimize power.

Thanks a lot once again for the information provided...
 
Hello Pietro
I can give you a couple is suggestions. First, you need some good meters. Here's a link to a clamp meter I bought, and recommend.
I first bought a very expensive, American-made Fluke meter used off of EBay. I subsequently found this one and tested them side by side. I found that the "affordable" Chinese UniT meter was accurate to within 1% of the Fluke readings, which I thought was outstanding. Here are a few things you need to know about clamp meters. They come in a wide variety of configurations, with the price going up as capabilities go up. The least expensive meters read AC amperage only. Higher in price are those that read both AC and DC. Another step up is AC, DC, and Inrush. Inrush is the measurement of the max amp draw in the first fraction of a second at startup. A regular AC clamp will not be fast enough to give an accurate reading of inrush. So, for a solar installation, the one I picked for you can do AC, DC, and inrush. Note that these meters can also be used like a regular voltmeter, ohmmeter, ect. so it's everything you need for setting up a solar system.

Here's an important detail about clamps. For the clamp to work, you clamp the jaws around only one of the electrical wires feeding the equipment. You cann't measure the ampeage of your toaster for example if you clamp around the power cord. The power cord has two wires, hot and neutral bonded next to each other, and the two wires cancel each other out and the net reading is zero. You could make it work by taking a razor blade and splitting the wire lengthwise and separating the two, but who wants to do that? What you can do is also buy an AC splitter like this...
Now this one has an American style plug. What country will be you using this in? I assume it might be different.

Here is a link to my inverter, which produces both 120 and 240VAC. Note that I bought the NA version, which means "North American"
They also market an 230VAC European version. This brand inverter can be paralleled to increase power. Look through the various models of inverters on their website. Most of the Schneider inverters can be paralleled to increase power output. So can Outbacks. You'll just have to select the one appropriate for your needs.

The one thing I can tell you about these systems though is you can't use the word "affordable". I have a big solar system with 4500W of panels and the XW+6848 inverter. The whole system including batteries was >15k USD, and I'm still upgrading it. You can put something together that's going to work, but it ain't gonna be cheap!
 
I was looking at Beckhoff PLCs as well as AutomationDirect BRX series, but i found that company extremely difficult to deal with.

I use a LOT of the automation Direct CLICK PLC units. I have had no issues dealing with them. The CLICK units are super easy to work with, I could guide you if you would need. I'm actually strongly suggesting making some videos on them.
 
So many questions...

Since you are using a 'Well Pump', what are other pumps doing?
My well pumps make plenty of pressure to fill pressure (bladder) tanks to maintain pressure on the RO filters, so no need for extra pumps at all.
A pressure tank is the time tested way to provide pressure and NOT run the pump every time there is water used on the system.

This is a well pressure switch with High/Low pressure set points,
The tank gets pushed to maximum switch set point, and 30-40 pounds lower pressure (as water is used) the low pressure set point closes the switch and the pump runs again...
No continuous pump, only one part time pump needed.

If it's a well, is it a deep well (into rock) or is it a shallow surface water well?
My deep wells provide 'Hard', but perfectly acceptable potable water.
For about $30 the county extension agent can have your water tested for minerals, contamination & pathogens specific to your location in the USA.
My extension agent always says he wishes his city water was as clean as my limestone well water...

Minerals *Can* be an issue, as well as chemicals & pathogens, but 95% of hard rock deep wells with casings are safe (potable).

DC is more efficient than AC.
No inverter needed, so no inverter losses (15-35% depending on if you have batteries or not).
DC would also let you manually connect panels to pump in the event of problems with charge controller, batteries, etc (emergency redundancy).
 
Back
Top