ArrayON variable with stand-alone project

[julmou]

Good morning,

There is something that I find tricky to evaluate with stand alone projects, and that is highly influential on the entirety of the system. I noticed that with "oversized" PV array (PV modules cheaper so most systems tend to decrease cost on battery by having more kWp which would provide current even under cloudy conditions), there's more wear on the battery because it is actually used more. The reason being that it reaches the max. SOC more often -> and we have the charging thresholds at 90% (triggered OFF -> array disconnected from battery) and at 75% (triggering back ON -> array reconnected) -> so what ends up happening is that the PV array is disconnected often, thus the load is often working on battery only, without direct use from the PV array -> therefore, even though we have plenty of energy available, the battery wears down more quickly. See below snapshot of the Charging Thresholds and Hourly Graphs showing the PV array disconnecting:

NB: this is with direct coupling

Anyway, where I am getting with this, is that it would be great to have a variable helping us estimate the importance of this phenomenon, such as how many hours the PV array is actually ON every day -> the ArrayON variable (Duration of the PV production of the array) should be perfect!

Unfortunately, ArrayON always have the same value as long as the sun is out -> which is where I think it's wrong and I can only suggest to re-assess this variable for stand-alone systems -> when the Array Current is 0 (not delivering any current to the system, PV array disconnected at the charge controller level), then ArrayON should also be 0

ArrayON in "normal behaviour" (charges battery and provides direct energy to the load):

ArrayON when the PVarray is disconnected -> still ON, even when the array is not producing anything

 

Thanks so much for reading this (a bit longer that I intended it to be sorry). Let me know what you think.

Julien

[André Mermoud]

You are right: it is not very logical that the battery is used when there is sufficien PV power for feeding the user needs.

However if you think about the real system in operation: when the battery is full, it has to be disconnected.

This means that the PV array is directly connected to the user's circuit, through the Power converter.  If the user's needs are lower than the PV production, how does the MPPT-DC converter behave ?  

In a Grid-connected system, the Power limitation (which is a fixed value) is achieved by a displacement of the operating point on the I/V curve. A MPPT-DC converter should be able to detect the output needs level;  probably an overpower (with respect to user)  will increase the output voltage. The device should be able to detect this voltage increase and displace its operating point on the I/V curve accordingly, Which MPPT converters are equipped with such a control system ?  I don't know !

When I wrote this simulation behaviour (25 years ago! ) I was pretty sure that such a control doesn't exist. This is now perhaps possible  (this should be mentioned on the datasheets), and perhaps the simulation should be able to treat this case.   

 

 

 

[julmou]

Quote

However if you think about the real system in operation: when the battery is full, it has to be disconnected.

I don't argue that at all. I just think that, when this is the case (array disconnected because battery full), then the variable ArrayON (Duration of the PV production of the array) should also reflect that.

At the moment ArrayON just counts the hours as long as the sun is out. I think ArrayON should not count the hours when the array is disconnected and not producing any current and energy (as I tried to illustrate above -> i.e. when the array current or when the effective production of the array is zero).

Also, I haven't tested with MPPT or DC-DC converter yet. This is just with Direct Coupling.

[julmou]

I'm not sure if I was clear with this suggestion.

I just think the ArrayON variable (Duration when the PV array is producing) would be a really good indicator to give us an idea of how much the PV array is disconnected during a year (when battery reaches full capacity).

By comparing different simulations, we would be able to compare in which configurations the PV array is the least disconnected (so more direct use).

But at the moment, the ArrayON variable is still counting even when the "Array Current" and "Effective energy at the output of the array" drop to zero (in practice, PV array not producing anymore, so ArrayON should stop counting).

[julmou]

Has the ArrayON variable been corrected since then?

This variable, in hours, indicate how long the array is on.

However, with stand-alone projects, even when the array is switched off (because max load SOC trigger has been reached) and as the array is effectively off ("Array Current" and "Effective energy at the output of the array" both at zero), the ArrayON keeps counting the hours as if the array is still ON.

[julmou]

Any thoughts on this ArrayON variable that keeps counting the hours even when the array is switched off?