Off-grid well pump, lowest energy consumption: variable speed vs fixed speed?

[sharpener]

I suppose it would. We inherited a system without one and it has never bothered us, the pump is pretty noisy (?cavitation?) but is in the garage 20m away, I would not want it in the house without some kind of NR treatment.

 

In the extreme case my grandfather ran the downhole well pump to fill it up the tank in the loft only once a day, as a child I was fascinated by the cat-and-mouse indicator under the eaves which showed the tank level. There was also an automatic level switch but he did not use it as he said the frequent starts brought up a lot of sand as well.

 

We have rainwater so sand is not a problem but the pH is, have tried various things but never quite cracked it, will make sure to fill up the HP with mains water which is slightly alkaline. Vaillant recommend pH 8.2 min and I bet most installers never check this.

 

On 06/06/2024 at 18:34, Hastings said:

I ran a test to eliminate the suction piping entirely by putting the pump in the supply tank as a submersible, it still couldn't achieve 2 bar pressure (hence the conclusion that it was faulty).

 

I now see the OP has already done this test, with no resistance at all on the suction side it sounds to me like the pump is defo stuffed whatever the supplier says. And also too big for the inverter to handle. Truly the worst of all possible worlds ((c) Voltaire).

[JohnMo]

6 minutes ago, sharpener said:

pump is pretty noisy (?cavitation?)

Possibly caused by your longish uphill suction line. The pump is starting to pull a vacuum and this causes cavitation. You need to look at on the data sheet, at the NPSH(r) or net pressure suction head (required). This is amount of upstream pressure the pump needs to stop cavitation. Ideally a neg value would be good, then the pump can pull the water in uphill and with pipe/strainer pressure losses be away from cavitation.

[sharpener]

1 hour ago, JohnMo said:

Possibly caused by your longish uphill suction line.

 

Maybe, I had thought it might be, certainly doesn't sound very nice. But there is no info on the data sheet about this, it is pretty sketchy. No signs of erosion though when I had it apart after several years' use to discover the cracked shroud.

 

The info about the reduction in flow vs suction head copied above is actually taken from the  data sheet for an earlier ST jet pump, now discontinued, but I imagine they are all pretty similar. That model was supposed to cope with a suction lift of 7m (presumably without cavitation), and at low flows the pressure drop we need to add to take a/c of flow resistance will not be much (certainly compared with 20m of underground delivery pipe).

 

The DAB Jetcom 62 is a bit better and will cope with a 9m suction head but has a somewhat higher max power consumption, 850W. The small Clarke is 800W, so the ST Jet 55-45 at 650W is the best of the three if max power consumption is the chief criterion.

 

 

[SteamyTea]

Regarding the inverter, can you fit a local battery and small inverter to the pump. That way you can take the load of the main one. Will be some efficiency losses, but better than knocking the power out.

[Hastings]

@sharpener Thanks for taking so much trouble to help - it really is appreciated. You have educated me and put me onto some new thinking.

 

I didn't know about the inverter rating for pumps advice from Victron. My 4kW Victron is therefore strictly too small for a 1.2kW pump, although I wonder if Victron are talking about fixed speed pumps that have big startup spikes? I will look into that.

 

From your experience I think I would rather go with the DAB pumps as the parts are all replaceable (as with the Steelpumps pump I have).

 

Yes, I have tried the existing pump as a submersible and found it still to be unable to achieve more than 1.5bar at zero flow. The test setup was with just a 3m length of output pipe from the tank (including the same small pressure vessel specified by the manufacturer) to a stop valve. This result is what decided for me that the pump or controller was faulty and so I sent it back for repairs. But it was deemed not faulty and was reaching up to to 8 bar in their workshop (and I believe them because they have already sold me a replacement pump at wholesale price when the original pump was destroyed by gravel in the suction line).

I am now suspecting that my battery bank and/or inverter could be the problem. My uncle's motorhome had trouble with sensitive mains equipment that wouldn't work and the fault he traced eventually to poor state of the battery powering the inverter. My house battery bank, although only 3 years old (another long story) I am anyway having to replace (with exactly what you have) this autumn, so it makes sense probably to do nothing about the pump until the new b/bank is in place.

 

I have one section of spiral reinforced pipe(from a floating intake) in the tank and the rest, 25m underground to the house, is standard rainwater MDPE (black with a green stripe) which is probably not intended for suction lines but I think is strong enough to hold up as it is strong enough to deploy underground without any protection from compression forces there. It is 32mm but the pipe inserts at the joints restrict it down to 22mm which isn't great. I might try taking those inserts out.

Edited June 24 by Hastings
Added info

[sharpener]

4 hours ago, Hastings said:

Yes, I have tried the existing pump as a submersible and found it still to be unable to achieve more than 1.5bar at zero flow. The test setup was with just a 3m length of output pipe from the tank (including the same small pressure vessel specified by the manufacturer) to a stop valve. This result is what decided for me that the pump or controller was faulty and so I sent it back for repairs. But it was deemed not faulty and was reaching up to to 8 bar in their workshop (and I believe them because they have already sold me a replacement pump at wholesale price when the original pump was destroyed by gravel in the suction line).

I am now suspecting that my battery bank and/or inverter could be the problem.

 

Sounds really strange. What exact model/type/size of Steelpump is it? I see they have quite a range including jet pumps, new to me but look quite impressive. Then I can look at the data sheet for further inspiration.

 

Pipework sounds OK to me, much of my suction pipe is 22mm Polypipe, again sufficiently rigid not to collapse.

 

Did you check the voltage at the terminals when you were doing the test? Victron inverters are robust and well regarded, they put out a decent sine wave so I would be a bit surprised if that is the problem. Also you can set various alarms inc low battery and overload, and look at the history using VRM to investigate problems, does this reveal anything e.g. running the pump pulls the DC voltage down unexpectedly? If your pump has a VSD that would almost certainly act as a soft start lessening the impact on the inverter somewhat. How does it cope with other heavy loads e.g. cooker or fan heater?

 

My pump in contrast is a simple capacitor-run type with direct-on-line starting so I am not surprised it makes the lights dip a bit but as the tests showed (I needed to check it anyway) it will start reliably off-grid without crashing.

,

Victron advice is not specific to pumps, applies to all motors, see for example this discussion, if you follow the link there is a video where at about 12'15" in Johannes Boonstra explains the limitations for starting motors. I keep an eye on this forum and there have been various queries about motor starting, I don't recall anything which might shed any further light on your problem but you could always post your own query there.

 

I have got 3 x Pylontec 3.55 kWh Force-L2 modules giving me 10.65 kWh nom capacity. Two modules was defo too small so I added another after their first winter. They are a bit smarter looking for a domestic install than the usual 3000 series and the wiring is much neater but I managed to get them for about the same price. They are the same electrically, the only problem peculiar to all Pylontecs is they are only 15 cells and they will not supply full rated current when below about 20% SoC as the internal resistance means this takes the voltage below the Victron trip threshold. So I have written a NodeRed flow to taper off the max inverter power before this happens (see my post Pylontech batteries will not cook the breakfast).

[Hastings]

17 hours ago, sharpener said:

What exact model/type/size of Steelpump is it?

X-JE120BT 3-phase Evolution series pump

E-Power e-MM/MT 230V controller/inverter

Yes, it is a soft start.No spike - takes a few seconds at startup to go from 0-1200W.

 

Other high power kit in the house: air-fryer, small kettle (1.7kW each), shower room heater 1kW work fine - battery DC voltage drops temporarily as expected, followed by rapid recovery. Also have the 170L Edel ASHP hot water tank (300-400W) working fine. Battery bank: 8x12V 210Ah (C20) Deep Cycle AGM Rolls.

 

It'll take me a while to check through your other links and pointers (thanks).

[SteamyTea]

3 hours ago, Hastings said:

Other high power kit in the house: air-fryer, small kettle (1.7kW each), shower room heater 1kW work fine

They are all resistance loads, the water pump is an inductive loads. The starting characteristics are very different.

[sharpener]

19 minutes ago, SteamyTea said:

They are all resistance loads, the water pump is an inductive loads. The starting characteristics are very different.

 

Yes. I thought it would be useful to compare though.

 

Failed to realise it was a 3-phase pump motor driven by 1 to 3 phase inverter drive. Am currently reading through the manual and have already come across a couple of mistranslations so could take a while! It's possible I suppose that the one inverter does not like feeding the other or vice versa.

[SteamyTea]

49 minutes ago, sharpener said:

Failed to realise it was a 3-phase pump motor driven by 1 to 3 phase inverter

I seem to remember (been 18 years since I seriously studied electronics and I was crap at it) that back electro motive forces can play havoc with the harmonics.

Edited June 25 by SteamyTea

[sharpener]

Meanwhile I see the Steelpumps X-JE120BT is in fact a jet pump from its shape and as mentioned in the text.

 

Performance of the 800W version seems to be very similar to the DAB Jet 62 upthread.

 

I  would question the supplier having found @Hastings's 1200W one produces 8 bar on bench test though, because the mfrs data sheet shows it only going up to 5 bar when stalled (and that seems typical of the other jet pumps I have looked at):

 

[Hastings]

17 hours ago, SteamyTea said:

I seem to remember (been 18 years since I seriously studied electronics and I was crap at it) that back electro motive forces can play havoc with the harmonics.

The supplier said I couldn't extend the connecting cable between motor and controller much due to an issue with "harmonics".

The pump controller manual "recommends the addition of the Long Connection Adapter for cable lengths over 20m, obligatory for lengths over 100m, to reduce overvoltages created by using long cables between pump and controller".

 

[Hastings]

18 hours ago, SteamyTea said:

They are all resistance loads, the water pump is an inductive loads. The starting characteristics are very different.

And so an induction kitchen hob (not that I have one but I have been seriously considering it) is also not a resistive load but inductive as the name suggests?

[Hastings]

17 hours ago, sharpener said:

I  would question the supplier having found @Hastings's 1200W one produces 8 bar on bench test

I may have mis-remembered this. I am not very technically minded and pretty bad at storing numbers in my head during phone calls - it was a few weeks ago.

[Hastings]

I have a cheap single phase 750W dirty water pond pump. It might be a useful test of the power supply to see what pressure I can achieve with that. I'll need to buy a pressure gauge.

[Hastings]

22 hours ago, Hastings said:

the 170L Edel ASHP hot water tank (300-400W) working fine.

@SteamyTea This I think is a motor based device driving a small fan and compressor? So also an inductive load? I guess too small a load to have significance re. the issue under discussion.

[SteamyTea]

22 minutes ago, Hastings said:

I think is a motor based device driving a small fan and compressor? So also an inductive load? I guess too small a load to have significance re. the issue under discussion.

Possibly.

Hard to tell without testing.

Some inverters have a minimum load requirement as well i.e. 200W.

Though I don't think this is your problem.

[sharpener]

Here are a few further thoughts and possible lines of enquiry

 

1. Inverter drive - Compatibility with pump etc

Has this combo been tested/approved by Steelpumps? What do there Technical Sales ppl say about it?

 

On 25/06/2024 at 12:37, Hastings said:

X-JE120BT 3-phase Evolution series pump

 

Italian 3 phase is 230V phase to phase. UK 3-phase is 415V phase to phase. Steelpumps therefore make two different versions of the 3-phase models, have you got the right one for the VSD (I think -230 not -400), you didn't give the suffix?

I see the inverter instructions say
"For having correct behaviours it is essential to choose a
pump with slope characteristic curve (see fig.), usually
multiimpeller pumps; this kind of pump permits the
Epower to pilots pump at variables speed.
The head and capacity of the pump must correct for
request of the plant."

So if it has to be a convex characteristic and the pump has a straight line as upthread maybe the inverter s/w can't find a proper working point on the curve. N.b. this is not a multi-impeller/multistage pump. (The English "three-stage pump" on p.9 is a mistranslation of "pompa trifase").

The inverter drive has a lot of programmable functions. Maybe one of the limits (?voltage ?frequency ?pressure) is set inappropriately. There is a factory reset that can be applied to go back to the default settings, p.33 code 76.

 

If you have to re-locate the/a pump to the tank site I would ditch the VSD, as the complication of needing the Long Connection Adapter if you do not also move the drive and also needing to run three phase cable is a PITA.

 

2. Choice of pump

 

Looks OK and is suggested for this application. But the 1200W version would probably be too big for the Victron inverter if you did not have a soft start. Inductive loads such as motors are not in themselves a limitation, the inverter is specified for a power factor of 0.8 anyway. I would guess the 800W single phase version would be OK. Maybe you could keep the pump but replace the motor with the 800W single phase one?

 

Should be capable of a reasonable suction lift, they don't show the figures but DAB and ST will do 9 and IIRC 7 m respectively. The effect from the graphs seems to be to restrict the max flow, but not to reduce the delivery pressure at lower flows.

 

What do Steelpumps technical say about using it with such a long suction line?

 

Is there a pinhole air leak in the suction line somewhere? If you leave it overnight is the foot valve or NRV sufficient to prevent any drop in level?

 

As a sanity check this online calculator says 25m of 28mm plastic pipe will require 105mbar of head for a flow of 20 l/min. So will add 1.05m to the suction lift which is reasonable. Velocity of 0.8 m/s is also reasonable. For the fittings assume 1m extra length per elbow, but it will not make much difference.

 

3. Is the pump broken?

 

On 24/06/2024 at 13:44, Hastings said:

Yes, I have tried the existing pump as a submersible and found it still to be unable to achieve more than 1.5bar at zero flow. The test setup was with just a 3m length of output pipe from the tank (including the same small pressure vessel specified by the manufacturer) to a stop valve. This result is what decided for me that the pump or controller was faulty and so I sent it back for repairs. But it was deemed not faulty and was reaching up to to 8 bar in their workshop (and I believe them because they have already sold me a replacement pump at wholesale price when the original pump was destroyed by gravel in the suction line).

 

As above I would query the reported 8 bar on test. It sounds as though the pump is faulty or is not getting enough power from the VSD for some reason.

Does it sound as though it is operating at the right speed?

Can you try connecting it directly to a 3-phase mains supply somewhere? N.b. do not do this if it is a 230V 3-phase motor (see above).

Has one of the windings gone open circuit (you should read the same resistance between any two of the three terminals)?

Have you monitored the terminal voltage and frequency when being driven by the VSD? Does it provide the same voltage to all three phases? You can read the frequency off the VSD display or infer it using a simple RC circuit and a voltmeter.

 

4. General thoughts

 

VSDs are of most benefit where you have a pump that has to operate continuously but with varying loads/flow rate. I don't think the savings will be significant in yr domestic situation.

 

The soft start is beneficial in cutting the initial power drain but the slow startup will mean without an accumulator there will be a corresponding delay in providing the flow compared with a simple pressure/flow switch (which seems to work well enough for me without crashing the battery inverter).

 

HTH

 

 

 

[SteamyTea]

3 minutes ago, sharpener said:

the inverter is specified for a power factor of 0.8 anyway

Which should cover almost all inductive loads.

What happens if it does happen to go out of range, are there any error codes?

[sharpener]

1 hour ago, SteamyTea said:

What happens if it does happen to go out of range, are there any error codes?

 

Not that are specific to this situation. In the present situation it might struggle to start a 1200W pump on its own, but there is a soft-start VSD between the pump and the Victron setup anyway.

 

If as the OP believes the battery/inverter is struggling I would expect to see a low battery or overload warning but he doesn't. If there was a loose connection somewhere that would probably throw an error too. @Hastings what Battery Management System do you have with your Rolls LA batteries? I am puzzled you say the pump crashes the system unless the batteries are fully charged, I would expect (i) the soft start provided by the VSD to obviate this problem and (ii) it would show up as a low battery error message. Are all the battery connections fully tightened?

 

The Victron inverters have according to this video a 100% overload capability for 2 mins, they are widely used for off-grid pumps machinery etc and there have been a few queries on their forum about induction hobs and really large motors. The video also recommends allowing 100% margin for reactive loads with a PF 0.7 to 0.8, but that is less restrictive than the advice to allow for 6x startup current for a single-phase motor.

 

As upthread I tried with the kettle etc already drawing 2.4kW and it will still start my 650W pump with the briefest of flicker. I looked to see if my bench grinder would be more of a test but it is only 260W and I do not have anything bigger. (The vacuum cleaner is 1200W but IIRC it is a series-wound AC/DC type motor so not yr typical induction motor like the pump. Ditto the lawn scarifier.)

Edited June 26 by sharpener

[dpmiller]

I'd caution that a multiphase motor on a VFD might very well be running on a reduced voltage by design

[sharpener]

4 hours ago, dpmiller said:

I'd caution that a multiphase motor on a VFD might very well be running on a reduced voltage by design

 

Yes but if the output is not reaching the target pressure presumably the VFD will increase the voltage/frequency in an attempt to achieve it, subject to the limits that have been set.

[sharpener]

OK @Hastings I think I have an answer for you but it is quite technical.

 

The E-Power e-MM/MT Variable Speed Drive probably uses a simple bridge rectifier to provide an internal DC supply as per the description here. The rectifier will conduct over a very small angle near the top of the input sine waveform and this will result in it taking short duration heavy current pulses from the mains input. I suspect given its compact size and economical price this VSD does not incorporate the proper filtering on the input necessary to avoid this effect.

 

These heavy current pulses are something the Victron Multi inverter probably does not like, and in consequence it may not be able to power the VSD properly especially at low battery SoC. I have posted a query about this on the Victron forum, however the replies so far are about using industrial VFDs on a Multi RS Solar (which is anyway a different type of inverter not using a toroidal mains transformer) so may not be applicable to your case.

 

Are you able to take the pump/VSD combo to somewhere there is a grid supply to see if it then works? Did your supplier sell you them specifically for use with the Victron inverter? In which case did you get a guarantee that it will work OK?

 

Long term I think you would be fine with Steelpumps' smallest jet pump with a simple pressure/flow switch, but ideally you would want to check with them that the long suction line is OK before committing.

 

HTH

[Hastings]

@sharpener Wow. I am grateful, somewhat lost for words and humbled by the expertise, time and effort you have put in to this. Thank you.

 

I had already been thinking about the logistics of taking the equipment to a friend or neighbour's house to test. Now it looks very much like a worthwhile undertaking.

 

Yesterday I tried running the generator into the system while pumping water but it made no difference. The generator is only 1kW, its function being to charge the battery bank when solar is insufficient, so I cannot run the pump off it directly. The SoC of the battery bank doesn't vary the behaviour of the pump at all other than stopping it entirely when the DC voltage drops below the Victron shutdown limit.

 

So I think you have very likely found the answer but frustratingly I can't do the phone-a-friend and house mains test as I've just left site for the next 2 weeks.

 

No, I have no special

arrangements with the supplier to allow for returning the equipment but I'm a dab hand at eBay, learnt from disposing of outdated photography equipment during a career as a cinematographer.

Edited June 27 by Hastings
typo

[Hastings]

On 26/06/2024 at 15:06, sharpener said:

what Battery Management System do you have with your Rolls LA batteries? I am puzzled you say the pump crashes the system unless the batteries are fully charged, I would expect (i) the soft start provided by the VSD to obviate this problem and (ii) it would show up as a low battery error message.

The batteries are managed directly by the 48V 5kVA Victron EasySolar (Multiplus and Solar charger in one unit). When the pump is pulling full 1200W and battery bank in good SoC the voltage drops from about 52V to 50V or so.