TSW3224 and deep well pump problem

System · October 2011

Hello folks,

I finally got to hook up my Apollo Solar TSW3224 to my house panel (with Interloc kit).
It succesfully powered lights, outlets, TV, computers and refrigerator.
VA readings were in the 800 to 900 VA. Voltge readings were 239 V.
Then I turned off all panel CB's.
Then I turned on the CB to the 1 hp, 240 V, deep well submersible pump. In other words no other loads except the pump.
The voltage reading drop to 115 V.
The pressure gage in the tank remained the same. It seemed the pump was not running. But there were current readings on the TSW3224.
After about a couple of minutes I decided to turn off the CB, concerned I might be burning my pump motor.
I put the panel back to grid power and the pump perfomed well as usual.
I put the panel again on the TSW3224 and the same problem occured.
Apollo Solar has a brochure that says the TSW3224 can start a 3 HP motor, and has a 7 second surge of 200% 6400 VA.

The battery bank is:
6 pairs of 12 V, 100 AH AGM. Each pair in series, and then the pairs are in parallel for a 24 V, 600 AH system.
Batteries have been charged to the last stage by the TSW3224 in charger mode.

Can any body tell me whats wrong here?
Is the TSW3224 undersized for a 1 hp 240V motor?
If it is, will a second TSW3224 hooked up in parallel with the first, be enough to run the pump?

Thanks in advance for any input.

Cariboocoot · October 2011

Re: TSW3224 and deep well pump problem

Welcome to the forum.

I'm going to suggest you check the DC wiring to the inverter. Six parallel battery strings is pretty difficult to keep even current flow through. This should be wired with equal length runs to positive and negative bus bars or at least common junction points. The other issue would be the wire size itself.

My concern here is that there could be Voltage drop on the DC side when asked to supply such a heavy start load. Although water pumps are hard to predict, having 6 kW surge for 7 seconds should start a 1 HP pump.

You can check the DC Voltage at the inverter with a DMM both with no load and when you try to start the pump. If you have a DC clamp-on Ammeter you can check the current flow as well.

A drop of AC output from 239 to 115 is very, very odd. Are you sure the output is wired correctly? I'm not familiar with the Apollo inverter; is it a true 240 VAC output or is there an autotansformer involved here?

I'm probably missing something obvious here; it's been one of those days.

mikeo · October 2011

Re: TSW3224 and deep well pump problem

Then I turned on the CB to the 1 hp, 240 V, deep well submersible pump. In other words no other loads except the pump.

on some thread here I posted a link to a pump manufacture that specified start loads for various pump motors. I suspect that your 24 volt wiring is way undersized or you have a bad connection to battery or panel. Your starting load is probably in the order of 70 to 100 or more amps. A 1/2 hp 230 volt submersible requires around 33 amps minimum to start.

waynefromnscanada · October 2011

Re: TSW3224 and deep well pump problem

And there is a difference in the ease, or difficulty of starting such pumps, depending on if the motor has a built in starter, and is supplied with just two wires + ground (for a total of 3 wires), or if it has a separate starter box with start capacitor, located above ground, usually in the house basement, and is supplied with 3 wires + ground (for a total of 4 wires) which run between said starter box and pump motor. The ones with built in starter are definitely harder to get going and I don't recommend them for off grid use for that reason.

Dave Angelini · October 2011

Re: TSW3224 and deep well pump problem

I would guess if you tried to start the motor with the internal charger running it might start. That would tell you not enough battery or it's wiring.

Never mind, the above only works with an XW with load support. Try it anyway!

System2 · October 2011

Re: TSW3224 and deep well pump problem

Hello Folks,

Thank you very much for responding.
I uploaded some pics to shed more light on my battery bank.
It is located in my crawl space, partially buried to increase probability of surviving a tornado.
I used 1/16" x 1" copper bus bars to connect the batteries together.
The posiive bus bar is in the foreground and the negative bus bar is near the wall.
I used c-clamps for several reasons: no hole drilling on the bus bar no decrease in flow area, saves time in fabricating and dont have to deal with precise alignement of terminals to holes during the install.
The c-clamps stay tight by using a cut piece of rubber hose acting like a lock washer and keeping the threads tight on their contact surfaces.
I use 2/0 stranded cable from bus bars to the TSW3224 (TSW = true sine wve) DC terminals located directly above inside the house. Each cable (+ and -) is approximatley 12 feet long.
In between the batteries in series is a 175 A fuse and a short piece of bus bar.
You might be wondering what is that thing in between the + bus bar and the cable. That is a 160A (continuous) relay. This relay allows me to disconnect the battery bank at the + bus bar. In between the relay and the + cable is a 175 A fuse.
As for the pump wiring, its a 3 wire + ground. Then there is a control box above ground with a capacitor for starting the pump. Distance from control box to pump below is 200 feet.
From the panel to the pump the wiring is gage 10. The total distance from panel to pump is 300 feet.

Below are responses to your comments:
--You can check the DC Voltage at the inverter with a DMM both with no load and when you try to start the pump. If you have a DC clamp-on Ammeter you can check the current flow as well.
--The TSW has an LCD readout shoiwng the DC V, VA and amps. The V drops gradually during the run from 26 to 25 to 24.8 and stays at 24.8.
---Are you sure the output is wired correctly?
---Yes the inverter powers all the other loads as mentioned in my OP. I even plugged in appliances in different branch circuits to make sure. I used gage 8 wiring (3 wire and ground) from inverter to panel (20 feet).
----I suspect that your 24 volt wiring is way undersized or you have a bad connection to battery or panel.
----I selected the bus bar size and cable to handle the anticipated load of the pump.
I got some extra bus bars I might just double them up and see if that solves the problem.

Thank you.

mike95490 · October 2011

Re: TSW3224 and deep well pump problem

I'm stunned

feel lucky that you have not had a clamp pop loose and light a fire !

C clamps will NOT give you a low resistance connection

7,000 watts (60A @ 120V) starting current will be 300A @ 24V Twice that if your pump is at 240V.

Use a low wattage incandescent lamp on the inverter, as you power up the pump, it would indicate faster than a meter, if the inverter is being starved. Or wire up 2 auto parking light bulbs in series, and clip them on to your inverter input terminals.

What's the wire run from the inverter - pump like, and how deep is the pump ? All that wire does not help either. I went 300' with #6, to keep my 240V pump happy.

halfcrazy · October 2011

Re: TSW3224 and deep well pump problem

I also would question the buss bar size 1/16" by 1" will not carry 300 amps I do not think? I also agree the C Clamps are a very bad idea.

How long are the 2/0 cables to the inverter? Those should really be 4/0 if they are more than 3-4 ft long to help with voltage drop as well. (Disregard I looked back 24ft total of 2/0 cable this will aggravate the issue as well they really probably need to be 4/0)

I am betting if you upgrade to a good buss bar and use 2/0 cables between batteries and buss bar with some 4/0 cables to the inverter it will start the pump with ease. Batteries cables are one of the week points they are so bloody expensive people tend to go under sized.

Dapdan · October 2011

Re: TSW3224 and deep well pump problem

Oiler,

I agree with HC. I don't think 1/16" copper will carry 300A. Besides that your setup looks scary with all those clamps and all. In fact I found some info for ampacity rating for copper buss bars here is the link:

http://www.copper.org/applications/busbar/ampacity/busbar_ampacities.html

Using this table 1/16" x 1" will carry approximately 137.5A which is way below 300A. So HC was right.

I have two Apollo inverters in the field and one at home. I had a similar problem with the tsw4048 one that is in my office. I had it running a 12000btu 240v AC. It ran it very well for a month or so and then it started to give an error message that was preceeded by a low voltage drop (similar to yours) when the AC was turned on. I called Herb at Apollo Solar to rule out the inverter. It was the load that was the problem. I AC had a bad capacitor and I changed it and it has now been working fine ever since. The moral of this story is:

1. Check with Apollo Solar technical team, they are good guys and I had good support from them.

2. More than likely it is your load that is posing the problem for the inverter. Is there a way you can check in rush to know exactly what this pump is demanding at start up when connected to the grid. When my AC was on overloading the inverter it was calling for more than 4000w as displayed by the inverter.

My final comments are for your batteries. Are they true deep cycle. It just seem strange that they would be deep cycle and have cylindrial post like those found on starting batteries. Usually even if it has those type of post it would be in combination with a stud. It appears to me these were designed for an atuomotive application that mostly uses the clamp on connection and not a deep cycle application that uses bolt on connections. I could be wrong.

Good Luck,

Damani

mikeo · October 2011

Re: TSW3224 and deep well pump problem

also would question the buss bar size 1/16" by 1" will not carry 300 amps I do not think? I also agree the C Clamps are a very bad idea.

A 1/16 x 1 inch copper is equal to #1 awg wire. It would take 3/16 thick by 1 inch to be equivalent to a 4/0 wire if I calculated the cross sectional area correctly. You are way under wired with your batteries.

waynefromnscanada · October 2011

Re: TSW3224 and deep well pump problem

mike90045 wrote: »

I'm stunned

feel lucky that you have not had a clamp pop loose and light a fire !

C clamps will NOT give you a low resistance connection

TOTALLY agree! I could not believe my eyes! These are VERY poor connections even if a clamp never ever comes loose. DEFINITELY NOT the way to go! No wonder your poor pump can't get started

mike95490 · October 2011

Re: TSW3224 and deep well pump problem

oiler897 wrote: »

...1 hp, 240 V, deep well submersible pump...

I'd expect that to pull 2,000 watts running, and 10KW when starting. You should consider a 48V system for that much power. (a lot to do with power factor)

That's 416A @ 24V (without inverter losses, which will add another 10% at least)
Only for a second or two, but it's enough to sag the DC voltage to crash the inverter.

in a 48V system, that would only be 208 amps....

solar_dave · October 2011

Re: TSW3224 and deep well pump problem

I also suspect the line contact on the posts provide a high resistance on the connections

System2 · October 2011

Re: TSW3224 and deep well pump problem

Thank you so much folks for the constructive comments and advice.
As a result I am going to redo my battery bank and connections.

Here is the battery specs, although it does not say it I think they are deep cycle.
http://www.altestore.com/store/Deep-Cycle-Batteries/Batteries-Sealed-Agm/Universal-Ub121000-12V-100Ah-20Hr-Sealed-Agm/p4163/

Here are the things I plan to do hopefully this weekend:
1) Get rid of the c-clamps and just drill holes thru the bus bars and use bolted connections.
2) Triple layer the bus bars on the + and - headers. That is put a couple more 1/16" bus bars on th existing headers making those header bus bars 3/16" thick.
3) In addition to the existing 2/0 cable from bus bars to the inverter. I'll also run a 4/0 cable from bus bar to inverter in parallel and same length.
4) Get rid of the 160 A relay.

After I implemented the above, Ill test it again with the deep well pump. If there is an improvement (like voltage is 195 instead 115) then I'll know I'm on the right track.
My next step then will probably be paralleling a second TSW3224 (a total of 12800 VA of surge for 2 seconds)

I appreciate all of your help.
Thank you.

BB. · October 2011

Re: TSW3224 and deep well pump problem

You can also check with a 1,500 watt AC heater and a DMM (digital multi meter) and look for voltage drop around your system.

Place a constant and large AC load on your system (say the 1,500 electric heater), then measure the voltage/voltage drop at each major point in your power system. If you have a DC current clamp meter, you can also measure the current in each paralleled battery string to see if they are sharing load equally. Note that the voltage drop measured for 1,500 watts should be roughly 2x larger for a 3,000 watt motor load, etc.

For example, if the battery reads 12.5 volts and the wire/bus reads 12.0 volts--then you have a 1/2 volt drop from battery to bus... And you can even set the meter to 2 volt or 0.200V/200mVolt scale and see if it is your connection or copper runs that are costing you power.

Same thing, measure voltage drop at the bus (say 12.0 volts) and at the input to the inverter (say 11.0 volts)--10.5 volts is cutoff for most inverters (2x for 24 volt system, 4x for 48 volt bank).

It is also possible that it is just your battery bank is not capable of the surge current either.

-Bill

Fullpower · October 2011

TSW3224 and deep well pump problem

Sixteenth inch thin copper flashing battery interconnects held on with C-clamps...
You have enough resistance right there in that dangerous combination to prevent start-up of a deep well pump.
You are indeed lucky to have not melted down any battery posts.
Your pictures show the scary-est battery cobbling up I have ever seen, anything you do at this point should be an improvement.
If I may make a recommendation, a minimum size for physically short series battery jumpers is 2/0 gauge cable, with minimum 5/16 through-bolted connections at each battery post, and any longer cables from battery to inverter would be better made from 4/0 cable.

techntrek · October 2011

Re: TSW3224 and deep well pump problem

3) In addition to the existing 2/0 cable from bus bars to the inverter. I'll also run a 4/0 cable from bus bar to inverter in parallel and same length.

Paralleling wires is bad, stick to one larger one.

ffsooo3 · November 2011

Re: TSW3224 and deep well pump problem

techntrek wrote: »

3) In addition to the existing 2/0 cable from bus bars to the inverter. I'll also run a 4/0 cable from bus bar to inverter in parallel and same length.

Paralleling wires is bad, stick to one larger one.

Could you explain or point me to a thread or site or anything that explains why parallel wiring is a bad idea? Thanks.

mike95490 · November 2011

Re: TSW3224 and deep well pump problem

Check out this artical about wireing on the diagonal

Has sketches and text that explains why

BB. · November 2011

Re: TSW3224 and deep well pump problem

From my point of view--the problems with parallel battery strings (some may be non-issues for you, others much more):

Paralleling Batteries is difficult because very minor changes in cable lengths, dirty connections, and even battery temperature/age can cause different strings to have different amount of current (instead of 4 strings each carrying 1/4 the current--one string may carry 1/2, another carry 1/3, and another carry 1/6, and one no load/charging current at all). I suggest a DC Current Clamp Meter (here is a cheap one that does work OK) so you can monitor the current sharing and look for problems before it damages your battery bank. It can be difficult to get batteries to share current well.
Paralleling can hid problems... If you have a bunch of parallel strings and one is open/high resistance. You will never notice the difference except if you need the energy (bad weather, failed generator, heavy starting load, etc.). And is is possible that that string has killed the rest of the batteries in the string because of no/under charging (again, clamp meter).
Battery cells fail--Sometimes shorted and sometimes open. A shorted cell can discharge, or even take down, the entire battery bank over a period of time (say winter, little solar, nobody there to monitor). Open cells will stop a string from charging/discharging.
Shorts happen... In cells, cable failure, tool falls on bank. You should have a fuse/breaker on each battery string output. Each time you add another string, it needs another (not cheap) fuse or breaker for that string. Nothing like watching the wiring turn red hot and nothing but an ax or bolt cutter to break the connection (even worse when you are not there).
If you have flooded cells, you have more battery cells to check/add water to. A 48 volt battery bank with 24 cells is one thing... A 4 string bank with 96 cells to check/water/clean can be a real pain.

So, my personal recommendation (this is from my experience with DC power systems in computers)--I would aim at one string of large AH capacity cells... And do no more than two or three parallel strings maximum.

With the large 2 volt (and 4 volt) cells available today, you can get pretty large ones that still can be moved around by one or two people and have a single ha AH string battery bank.

Also, some folks try to build (for example) very high AH 12 volt battery banks. With the charge controllers, inverters, etc. available today--people should move to 24 or 48 volt battery systems instead. This also reduces the current (charge controller and inverters) needed to operate. Getting over the 100-200 Amp DC power range requires a lot of expensive copper and breakers/fuse assemblies to do right. (100a*12v=1,200 watts; 100a*24v=2,400 watts; 100a*48v=4,800 watts).

Plus, for larger systems (higher current, longer cable runs), voltage drop becomes a nightmare... A 12 volt system can tolerate around a 1 volt drop (maximum). A 48 volt system can tolerate around a 4 volt drop maximum (from battery bank to AC inverters). Add that the current is 1/4 as much in the 48 vs 12 volt battery bus--it is much easier to move a whole bunch of power through your system.

Again, my opinion, there are many people with more than 2-3 parallel strings that are very happy with them. I just would avoid paralleling strings if it were my system. And sometimes, it is just not avoidable (can't move large cells/batteries, cannot obtain large AH cells, cost is too high, simply need a very large AH battery bank, etc.).

-Bill

TSW3224 and deep well pump problem

Comments

Categories