News Flash.

On another topic a few months ago, I was asked why I use such a mix of MCB trips from different manufacturers:

And here's the answer:

This is the right-most of the 5 MCBs in the first photo.

Unfortunately the MCB did not trip automatically,
and I had to move the green lever down (off) by hand, despite it being very hot.

I became aware of an issue when the set of inverters ceased operation because the one which was connected via this trip could no longer detect a battery.

When such errors occur with inverters in parallel, all of them disconnect their output.
Ie there is no system-redundancy when one inverter fails.

I am discussing the need for better resilience with a number of inverter manufacturers,
but it's more technically difficult that you might at first suppose.

I will supply a fault report to Taixi, who manufactured the 125A MCB above.
They may ask that it gets returned.
If not, then I'll open it up and post photos here.

@transparent A new luna battery would be more expensive (£s/kWh) but if I could pick up a second hand one at a good price, it would mean I wouldnt need to change any inverters initially which would save quite a bit.

I think the majority of the benefit I would get from a bigger battery would be in the increased amount I could store from an overnight cheap tariff so hooking up the 2.66 kW array would be nice but not essential. Also 10 kWh would be plenty although 15.5 kWh would obviously be nicer 🙂

Ive found out some numbers that I was recording for a different reason last winter and while they will only be approximate at least it will give me a baseline to compare this winter to. If they are similar then it looks like a saving of approx £250 per year is possible. This doesnt include any benefit from trading on unit prices with the grid.

I think that makes the ROI for a £2500+ investment borderline but if I could pick up a used Luna battery for a grand that would be much more attractive as would spending £3000 on a new battery and getting £1500 back by reselling the existing Huawei battery, BMS and inverter.

Im going to keep an open mind and scan ebay while I collect more data 🙂

So just to keep this thread updated, here is the latest months data. October has been mild and the potential savings of having a second battery are only marginally higher than September at £8.77.

This is assuming all of my peak usage could be subsituted for off peak usage if I got a bigger battery. At the moment this is true but it might not be when things get colder.

So a second update to incude Nov and Dec. The savings that would be made by adding a second 5kWh battery module would be £63 so far. If I could fit a big enough battery to stop any peak rate grid use then the saving would be £86.89

@transparent This business of SOC readout with Lipo4 batteries is so unhelpful. My boat setup has a Victron 200Ah battery which, when in commission is solar charged or from the alternator via one of their dedicated BMS . In the winter there is little solar and it only runs a camera onboard and I top it up with a charger. Last week the BMS shut it down though the BMV 702 showed 76% SOC. I believe the BMV 702 is not a coulometer and in a low current situation does not register SOC correctly but can find no warnings about this, how low is a low current? Does the SOC inaccuracy affect BESS setups where the batteries do not run full cycles and the errors mount up over time?

I agree @jancold  But there is no simple way to resolve the SoC issue.

Manufacturers are able to continue selling BMS units regardless of the 'accuracy' of the SoC algorithm they contain.

They tend to make that algorithm efficient at reporting when a battery is nearing full charge when large currents are flowing.
That's how most owners treat their batteries,
and that's when the BMS receives the most accurate data because it's balancing cells as they reach peak voltages.

So if a BMS is rated at 100A, and you're recharging at 30A or above, then the BMS will tend to use that to 'calibrate' itself for its SoC definition.

But if you then leave the battery discharging at only 1A or so, it's not going to be able to discern the difference between the energy taken by the load and the amount of internal discharge between cell plates.

Even an expensive coulometer is going to struggle with that concept.

There's nothing to prevent you having another, secondary SoC measuring system on your boat.
Peacefair have a PZEM017 with RS485 output which could allow you to create your own algorithm using an embedded processor (Raspberry Pi etc).
It can handle up to 300A, but is unique among their products in having no integral display.

The PZEM017 costs about £10 direct from China, incl shipping & VAT, and somewhat more in the UK.

They're a friendly Chinese company, and can respond to messages in English.
Laura sends out their Newsletters.
And you can send emails to Lucy Li at service02@peacefair.cn

I'd advise you keep the question brief, such as "Is the PZEM017 able to offer greater accuracy for State-of-Charge than the PZEM015 with its integral display?"
You don't want to be seen frittering away their time for a one-off purchase.
This is a company that sells crates of 100+ units at a time!

@bontwoody , thank you. I'm looking to do a similar off-peak storage (ASHP, but no Solar PV) and was interested to compare your figures with my calculations. Could I just check? - your December off-peak usage is 550kWh, presumably this supports about 150kWh of charging your 5kWh battery once every day (assuming the PV is not charging it) - does that mean your normal off-peak December consumption is 400kWh (i.e. 400/756 = 52%)? 

My calculations for my house, assuming the ASHP running continuously gives (on a cold day like today) a non-BESS assisted off-peak (7 hours E.ON night rate) of 31% of total. This means a battery of 30kWh to fully support a midwinter day. This does support all DHW heating during the day, which I can shift some to night hours, which will go some way to account for the difference we have in off/on peak usage.

Grateful for any comments, I am moving toward committing to getting on with this project. 

Posted by: @s_gatorator

↑

This means a battery of 30kWh to fully support a midwinter day.

On the Hello Topic, you said you are from a background of electronics engineering.

So if you'd like to grow your battery storage to 30kWh or more, would you be considering a self-build?

We would be willing to guide you through the process of course,
and I suspect you already have a fair amount of the tools required.

See what I posted here about a partial DIY-build, using steel enclosures with integral BMS units.

Posted by: @transparent

↑

So if you'd like to grow your battery storage to 30kWh or more, would you be considering a self-build?

We would be willing to guide you through the process of course,
and I suspect you already have a fair amount of the tools required.

Hi @Transparent ,

Yes, the whole battery storage idea was prompted by a chance conversation at a service station stop a few months ago, where I was encouraged to look at Fogstar's kits, which led me to here and the very helpful commentaries on that thread and some others , so thanks you for all those.

The kits certainly make the financing returns look better/acceptable. (With 30kWh of battery, an 8kW Sunsync inverter, self-build and install and paying an electrician for connection and signoff, I am expecting costs to be £6000)

I expect I'll have questions when I start. Yes, I have or can get hold of tools and test equipment (hand tools, clampmeter, multimeter, bench power supply). I suspect a torque wrench may be handy rather than winging it.

Let me pick up four points from what you've written @s_gatorator

1:  If you opt for Fogstar/Seplos enclosures you will need to consider how they will be arranged in 3D space.

To achieve the calculated 30kWh you would need two enclosures each with sixteen cells rated 305Ah
or three enclosures if you use the more common 280Ah cells.

Each enclosure will weigh about 110-120Kg when loaded.

If you have enough space to floor-mount them, then that's fine.
I've used Unistrutt-style steel racking components to wall-mount my batteries, thus raising them above floor-level.

Have a think what would happen if that area flooded, either from a pipe-burst or with the roof blown off.
Moving batteries at short notice when something goes wrong is not an option!

2: Compare the Fogstar LiFePO4 cells against those available direct from China.
Both suppliers the same A-grade cells manufactured by Eve, although Fogstar has opted for a different terminal design.

The shipping agent I've used in China is Xuba at Shenzhen.

You can contact Brian at Xuba via the Alibaba messaging system.
He may be able to give you different pricing and delivery times for stock that's already at a warehouse in Europe.

I can explain the shipping options and tax arrangements if you wish to know more.

3: I don't think you'll be able to run the heat-pump and the rest of the house using a single 8Kw Sunsynk inverter.

When an ASHP starts, there is a current requirement of around 30A until the compressor has brought both sides of the fluid loops up to operating pressure.
This is more than just a switch-on surge from the motor.

I've been using three 5kW Growatt inverters for the past couple of years.

The next stage of the trial will see these replaced with three 5.5kW Ecco inverters from Sunsynk.

System resilience is an issue.
You need to consider what options you have if one inverter develops a fault.

4: Torque wrench: To tighten copper busbars onto cell terminals you need one which offers low torque.

Tightening a nut beyond 8NM can strip the threads in the aluminium.

I use a Wera A6 torque-wrench, which allows me to first set it at 5NM, then re-tighten at 7NM, before finally going to 8NM.

You can sometimes pick up an A6 model cheaply when they get returned to a seller.
They are often purchased by mistake when the customer actually wanted high torque!

Thanks @transparent , some really useful review points - 

Posted by: @transparent

↑

If you have enough space to floor-mount them, then that's fine.

Yes, happily I do. Planning to put them in a 'garage' storage area a few m from the electricity meter mounted outside. The temperature stays above 5C when it's -2 outside so that's good for charging (I know I've read somewhere that the cells should not be charged below 5C and hence may need heaters in some installations).

I'll check the burst pipe scenario (good point - there is plumbing in the garage) but it may be sufficient to arrange for sufficient runoff under the door seal directly down the slope outside and use a Seplos-style battery enclosure on casters.

Posted by: @transparent

↑

Compare the Fogstar LiFePO4 cells against those available direct from China.
Both suppliers the same A-grade cells manufactured by Eve, although Fogstar has opted for a different terminal design.

Sure, I'll do that, thanks.

Posted by: @transparent

↑

I don't think you'll be able to run the heat-pump and the rest of the house using a single 8Kw Sunsynk inverter.

When an ASHP starts, there is a current requirement of around 30A until the compressor has brought both sides of the fluid loops up to operating pressure.
This is more than just a switch-on surge from the motor.

I'm planning to use the inverter in a grid-tied zero-export mode. I understand that if the battery cannot supply the full house load then the inverter will supplement from the grid:  

This is from P52 of the Sunsynk ECCO 8kW SG05LP1 User Manual

ZeroExport: The hybrid inverter will not only provide power to the backup load connected but also give
power to the home load connected. If PV power and battery power are insufficient, it will take grid energy
as a supplement. The hybrid inverter will not sell power to the grid. In this mode, a CT is needed. For the
installation method of the CT, please refer to the chapter "CT Connection". The external CT will detect power
flowing back to the grid and will reduce the power of the inverter only to supply the local load, charge the
battery, and home load. 

Am I missing something here?

Posted by: @transparent

↑

System resilience is an issue.
You need to consider what options you have if one inverter develops a fault.

If the inverter develops a fault then I assume that power will be supplied from the grid only.

Posted by: @transparent

↑

Torque wrench: To tighten copper busbars onto cell terminals you need one which offers low torque.

Thanks - Yep, will look out for one.

Posted by: @s_gatorator

↑

I understand that if the battery cannot supply the full house load then the inverter will supplement from the grid:  

Ah, we're both right about how the ASHP would be supported from a Sunsynk inverter, but we've made different initial assumptions

You're expecting to connect your heat-pump in the position marked as Home Loads.
Appliances connected there will indeed be able to draw current from the grid in the event of the inverter's export ceiling being maxed out.

I had assumed your ASHP would be connected as a Backup Load, on the upper of the two 240v ports in the diagram.
In that case it would continue to operate during a power outage...
... but the inverter must be able to provide 100% of demand.

You can, of course, change how it is connected at a later date.
Your opinion might be altered according to the frequency with which outages affect your location.