I just responded on TTT's suggestion - which I would also do to daily cycle between 10% and 20% from midnight and then wait for Solar to charge and assist during the day.

11 minutes ago, Ingo said:

I just responded on TTT's suggestion - which I would also do to daily cycle between 10% and 20% from midnight and then wait for Solar to charge and assist during the day.

I did not respond on your post, hoping someone else would, as I am not 100% sure of the result, as their are caveats.

Like Keeping batteries charged uses Eskom at night if the batts where used, to prepare for the next shedding.

 

I agree, if there is a caveat to this I would like to know as well.

Practical worst case scenario: If you had load-shedding until 10pm then your batteries might be at 50% I 'think' it will stay like that until 10am until the charge schedule kicks in but we all know the next load-shedding schedule could kick off at 8am again and then we are a bit, hmm.. you get the picture. If however there is an option to charge from Grid after 10pm to get it up to that min SOC (80%) again it would be nice. At 80% it would be too low to 'cycle' at 00:00 but it will be enough to carry the next load-shedding cycle until power comes back online or solar charges and assists.

Just for everyone's benefit, I chose 00:00 to start the cycle because my (EV) Geyser elements are managed to NOT come on during that time and that I will get a slow and steady discharge from the batteries for that 10% or 20% discharge time period.

1 hour ago, Ingo said:

I agree, if there is a caveat to this I would like to know as well.

Practical worst case scenario: If you had load-shedding until 10pm then your batteries might be at 50% I 'think' it will stay like that until 10am until the charge schedule kicks in but we all know the next load-shedding schedule could kick off at 8am again and then we are a bit, hmm.. you get the picture. If however there is an option to charge from Grid after 10pm to get it up to that min SOC (80%) again it would be nice. At 80% it would be too low to 'cycle' at 00:00 but it will be enough to carry the next load-shedding cycle until power comes back online or solar charges and assists.

Just for everyone's benefit, I chose 00:00 to start the cycle because my (EV) Geyser elements are managed to NOT come on during that time and that I will get a slow and steady discharge from the batteries for that 10% or 20% discharge time period.

can you pick a cycle time? sorry missed this somewhere?

 

Setting a charge window is not an option as this charges the battery from mains. battery should charge from solar..

 

Unless you can set a charge window to a 20% SOC and it will not charge or use battery if it is above 20%? Then one can set the battery to a 1h charge window but with a very low SOC?

 

To re-iterate.

• I dont want my battery to charge from mains.
• I want to supply solar power to my mains load
• i dont want to supply battery to my mains load
• I want to cycle my battery to my critical load at night.

2 hours ago, Ingo said:

So if I read this correctly, if you set the 'Optimised with Battery Life' and set the charging schedule to charge from 10:00 to 00:00 and a SOC of 80% then it will use the batteries from 00:00 until it reaches 80% SOC and then wait until 10:00 to charge from hopefully Solar power - or if clouded over, from Grid??

Yeah, you can use scheduled charging to prevent discharge during certain times, but at the moment no discharge is allowed during scheduled charging. This actually an intended use case.

If you know python and you can code a bit, here is a patch you can apply to systemcalc. What this does is it allows excess PV to power the loads if you're in a scheduled charge slot and the target SOC is already reached, but again it won't allow the battery to discharge. Scheduled charging always aims to END the charging slot at the target SOC or more. Never less.

 

Edited April 8, 20197 yr by plonkster

20 minutes ago, plonkster said:

Yeah, you can use scheduled charging to prevent discharge during certain times, but at the moment no discharge is allowed during scheduled charging. This actually an intended use case.

This might be a stupid question but I assume if you are in a scheduled charging state, without Solar EG. 8pm, that it would provide power from batteries during a grid failure. That means the 'no discharge is allowed during scheduled charging' statement does not apply?

23 minutes ago, plonkster said:

Yeah, you can use scheduled charging to prevent discharge during certain times, but at the moment no discharge is allowed during scheduled charging. This actually an intended use case.

If you know python and you can code a bit, here is a patch you can apply to systemcalc. What this does is it allows excess PV to power the loads if you're in a scheduled charge slot and the target SOC is already reached, but again it won't allow the battery to discharge. Scheduled charging always aims to END the charging slot at the target SOC or more. Never less.

 

SO are we saying that if I have a charging slot allocated then it wont discharge battery it will use mains? What is SOC of battery is above the SOC for charging. will it just not charge the battery and allow mains to supply?

Just now, Ingo said:

during a grid failure

If there is a grid failure, then the inverter switches to backup mode and none of the charge/discharge limits or scheduled charging stuff applies. Even discharge limits on the battery will be ignored (cause those depend on the loads, nothing the inverter can do other than switch off... which is a decision better left to the BMS).

2 minutes ago, JTK said:

SO are we saying that if I have a charging slot allocated then it wont discharge battery it will use mains? What is SOC of battery is above the SOC for charging. will it just not charge the battery and allow mains to supply?

Correct. It stops charging and uses the mains to power the loads. If there is PV, that charges the battery and when the battery becomes full the excess goes to the loads. That is what "charging" means. The Target SOC is there merely to prevent charging from the grid past a certain point, and this is there because very often a customer needs to charge his batteries to only 80% (for example) to get through breakfast, but not more than that otherwise the battery is too full after breakfast and he has nowhere to store the PV of the day.

1 hour ago, JTK said:

Unless you can set a charge window to a 20% SOC and it will not charge or use battery if it is above 20%? Then one can set the battery to a 1h charge window but with a very low SOC?

The way I see it,  you set your charge window with a low SOC, but a long ( daytime) window.

 

1 hour ago, JTK said:

I want to cycle my battery to my critical load at night.

This is on the load side, it would involve segregating your critical and non-critical loads.

I think you'd need to supply more detail here, as to the the system's response to SOC and grid conditions.

Your intentions might be quite nuanced.

2 hours ago, phil.g00 said:

The way I see it,  you set your charge window with a low SOC, but a long ( daytime) window.

Let me tell you the whole long tale. So first off, what is the purpose of scheduled charging?

Scheduled charging is for people with a TOU tariff, that is time of use. Power is literally so expensive in peak times that it makes financial sense to charge the battery from the grid in off-peak times, and to then discharge the battery during peak times. So scheduled charging is about charging from the grid. That is the first important part.

A typical use of scheduled charging is after midnight but before breakfast, because breakfast is traditionally peak hour. The trouble with this time slot is that quite often you don't want to fully charge the battery. Ideally you want the battery to be at a low SOC after breakfast, because you have PV coming in for the rest of the day and you need somewhere to put it. That means you don't necessarily want the battery to be FULL just before breakfast, you want just enough power in the battery to get through breakfast. This is why there is a feature to stop at a particular SOC.

One can of course calculate the charge rate or duration that would be needed to get the battery to a specific SOC and implement this solely using a start and end time, but that would not be very accurate. It would also depend on the SOC before charging starts, which may or may not be the same every day. Hence a feature specifically for stopping on a particular SOC. Charging (from the grid) stops whenever the SOC is reached or the scheduled window ends.

When the scheduled window ends, ESS goes back into its default "optimised mode", where it attempts to eliminate imports from the grid.

This raises the question: What do I do (as the implementer of said solution) if I reach the target SOC before the end of scheduled charge? One has to decide what the user would expect and what would be the least surprising. To me it seems obvious that if we're preparing for a high-tariff breakfast, we will schedule charge to end (and optimised mode to restart) just as the high tariff comes into effect (eg at 6AM). At this point in time, we want the battery to be AT the target SOC.

That means, if we reach the target SOC early, we have to stop charging but we also have to wait before starting to discharge. So that is exactly what it does at the moment.

HACK: You can therefore use scheduled charging to schedule periods of no discharge, eg in my home we stop discharging during the evening peak (supper/bath time), because I don't have TOU tariffs and I don't want to empty my batteries by 8PM and then have no reserve for a possible failure later in the night (which has happened in the past).

Then some more issues came up, mostly about what to do with PV (if you have any). If grid charging is scheduled for the middle of the day when there is also PV, what should you do? Below the target SOC it is easy: The grid and PV simply charge together. Once we reach the target SOC, the grid stops charging, but now the question becomes: Do I go into optimised mode, or not?

If I go into optimised mode, then there might not be enough PV to carry the loads (and I cannot predict it either), so going into optimised mode has the risk of discharging the batteries below the target SOC. We don't want to do that, so there is only one option left: Loads run from the grid, PV charges the battery until we're outside the time frame.

We do however turn on the same features as used by "Keep Batteries Charged" so that once the batteries are full the excess will automatically go to the loads, so if the batteries become full because of PV during scheduled charging, the surplus will go to the loads.

I am presently testing a feature where the Multi will go into optimised mode once the target SOC is reached while preventing discharge using power limiting. So far it's looking very good.

4 minutes ago, plonkster said:

We do however turn on the same features as used by "Keep Batteries Charged" so that once the batteries are full the excess will automatically go to the loads, so if the batteries become full because of PV during scheduled charging, the surplus will go to the loads.

So the weakness is, if I limited the charging current from PV ( because that's the max rate my batteries could handle, say), then the excess PV power would be curtailed, and grid power would be used for the loads until the batteries were charged?

43 minutes ago, phil.g00 said:

So the weakness is, if I limited the charging current from PV ( because that's the max rate my batteries could handle, say), then the excess PV power would be curtailed, and grid power would be used for the loads until the batteries were charged?

This is fixed in Venus 2.23. We had that issue with older versions. If the battery specified a really low value, it would sometimes curtail PV power.

So just to recap, so I have it down pat:

I set my scheduled charge to X% at a slow rate throughout daytime (the window).

If it was an overcast day (0% PV production), the grid would supply all loads and ensure me X% SOC to face into the night.

However if it was sunny, any DC coupled PV and the grid would share load & charge together until X% SOC, whereafter all PV production would be used either for charging up to 100% and for loads, with the Grid supplying any PV shortfalls in load peaks. And I would face the into the night with 100% charged batteries.

No... I'm having trouble understanding what you're understanding but it sounds wrong.

so I played around with all options.. no luck. I cant get to use the meter and not have my batteries discharge. I considered the option of "limit inverter power" but here is my concern...

 I limit say the power to 750w. This is more than enough for my loads on critical. However if the vacuum cleaner gets plugged in or my wife blows her hair, then what happens? Does it supplement from grid? What if grid is off? Does the "limit power" still count? Will the inverter "overload"? Also, if I limit inverter power... Does it limit which is fed back into mains load? That would be sad as I would waist all my excess MPPT generation...  Still trying to find the magic bullet. But yes. the charge schedule may have to be the option with me programming geyser heating in that slot... I dont like it though, to be honest. It seems like a band-aid to a sore tooth.. Not really fit for purpose....

13 minutes ago, JTK said:

so I played around with all options.. no luck. I cant get to use the meter and not have my batteries discharge. I considered the option of "limit inverter power" but here is my concern...

 I limit say the power to 750w. This is more than enough for my loads on critical. However if the vacuum cleaner gets plugged in or my wife blows her hair, then what happens? Does it supplement from grid? What if grid is off? Does the "limit power" still count? Will the inverter "overload"? Also, if I limit inverter power... Does it limit which is fed back into mains load? That would be sad as I would waist all my excess MPPT generation...  Still trying to find the magic bullet. But yes. the charge schedule may have to be the option with me programming geyser heating in that slot... I dont like it though, to be honest. It seems like a band-aid to a sore tooth.. Not really fit for purpose....

JTK, are you running a split board? thought the batteries can only output on the "essential out, and well the grid should be on the AC_IN which is on the opposite side, not reachable by the batteries.

G

1 hour ago, georgelza said:

JTK, are you running a split board? thought the batteries can only output on the "essential out, and well the grid should be on the AC_IN which is on the opposite side, not reachable by the batteries.

G

thats just the thing. when you ad the ET112 the inverter discharges the battery to the load that is attached before the AC-in... exactly not what I want

5 minutes ago, JTK said:

thats just the thing. when you ad the ET112 the inverter discharges the battery to the load that is attached before the AC-in... exactly not what I want

Post you VenusGX ESS settings here?

I ask again, is your system/DB split...

#1 Grid feed, CG, Geysers, non essential plugs => AC_IN

#2 essential load => Plugs and logs. => AC_OUT_1

It's only suppose to send battery to the "essential AC_Out" side, not the AC_IN (where the grid and the CG is).

G

13 minutes ago, georgelza said:

I ask again, is your system/DB split...

#1 Grid feed, CG, Geysers, non essential plugs => AC_IN

#2 essential load => Plugs and logs. => AC_OUT_1

It's only suppose to send battery to the "essential AC_Out" side, not the AC_IN (where the grid and the CG is).

G

yes. split DB. geyser  and non essential sits on AC1. Battery gets sent to this load in ESS mode with meter attached.

ps: meter is unplugged atm

25 minutes ago, georgelza said:

I ask again, is your system/DB split...

#1 Grid feed, CG, Geysers, non essential plugs => AC_IN

#2 essential load => Plugs and logs. => AC_OUT_1

It's only suppose to send battery to the "essential AC_Out" side, not the AC_IN (where the grid and the CG is).

G

In normal operation, the Multiplus-II generates power to both AC_IN and AC_OUT (they are directly connected by a set of relay contacts). It is only when mains fails that the AC_IN is disconnected from AC_OUT.

Assuming you are not trying to push power back to the Grid, the MP2 adjusts its power output to try and make the overall input power equal to the "Grid Setpoint" value (e.g. 40W). If you do not have any external current sensor, it does this using the internal current sense on AC_IN. In this case the MP2 will only use it's battery to supply power to the critical loads on AC_OUT2.

Adding an external current sensor has the advantage of changing the point where one wants to achieve the "Grid Setpoint" value.  By moving it "upstream" of the AC_IN connection allows the MP2 to supply power to more than just the critical loads. Any loads connected between the Current Sensor and MP2 will now also be able to be supplied from the battery (but not during power failures).

I hope this helps!

5 minutes ago, NigelL said:

In normal operation, the Multiplus-II generates power to both AC_IN and AC_OUT (they are directly connected by a set of relay contacts). It is only when mains fails that the AC_IN is disconnected from AC_OUT.

Assuming you are not trying to push power back to the Grid, the MP2 adjusts its power output to try and make the overall input power equal to the "Grid Setpoint" value (e.g. 40W). If you do not have any external current sensor, it does this using the internal current sense on AC_IN. In this case the MP2 will only use it's battery to supply power to the critical loads on AC_OUT2.

Adding an external current sensor has the advantage of changing the point where one wants to achieve the "Grid Setpoint" value.  By moving it "upstream" of the AC_IN connection allows the MP2 to supply power to more than just the critical loads. Any loads connected between the Current Sensor and MP2 will now also be able to be supplied from the battery (but not during power failures).

I hope this helps!

Thanks Nigel. This does align with my experience. But it just means the meter is a waist of money for me cause I wont use it :). No need for a meter on AC-out 1 and 2 as the MP knows this already. And I dont want battery power fed into AC-in.

I wanted to have a handle on my net mains watts. As in just use the ET112 as a monitor of current. Seems I wont get that option.

Edited April 9, 20197 yr by JTK

3 hours ago, JTK said:

However if the vacuum cleaner gets plugged in or my wife blows her hair, then what happens? Does it supplement from grid?

Yes, the rest comes from the grid.

3 hours ago, JTK said:

What if grid is off? Does the "limit power" still count?

The limit only applies when the Multi is running grid-tied. If the grid goes down, then all that stuff is out the window, then it acts like a normal UPS and it keeps the loads up.

So short answer: When there is no grid, the limit is ignored. It won't switch off, it will run the battery into the ground trying to power your loads (just like any other inverter would).

3 hours ago, JTK said:

Will the inverter "overload"?

When there is no grid, then the possibility of overloading the inverter is greater (of course). I've overloaded my Multi at least once by forgetting to turn off a heavy appliance before load-shedding. Also, if you run it above 80% for long periods it might overheat and shut down too. But to answer the question I think you are asking, will it "overload" because you exceeded the 750 limit set on the Venus device? No, it won't. That limit is only used when the grid is there.

3 hours ago, JTK said:

Does it limit which is fed back into mains load?

It limits how much power the inverter component feeds in, and that affects both input-side and output-side loads (the input and the output is just latched together, so it's really sort of the same thing). So to make that practical, if you set it to 750W, and you have 250W load on the output, it will feed max 500W to loads on the input side.

@plonkster On your last point. the question is. with the limit. will it still feed excess MPPT to the primary load side? Or will this also top out at the limit value?