high reactive power after solar install.

Status
Not open for further replies.
W

wwstrick

Member
Location
Antigua, Guatemala
I installed a 7600 W Solaredge inverter at a site that is a little unusual. They have one electric meter with four areas of consumption spread over a farm. It is about 1 mile from the residence. The solar is located at a point where a lot of energy is used to pump water. The service is 240/480V split phase and there is a transformer at each of the points where they use energy.The inverter is on the low side of one of the transformers. Before the inverter was installed they had registered on their bill 3100 kWh and 2870 kVARh. With the inverter producing 1250 kWh the electric bill after the install had 1638 kWh and 6280 kVARh. The savings for the kWh went to pay the excess kVARh. I can set the inverter to maintain an output of a set % of reactive power. I am considering setting it to 20% and watching the meter for a few days. Any ideas on why it would shoot up? Does the change to inverter output sound like a good idea?
 
vars went up because you have a very inductive load that the inverter is not reacting to. Complicated to explain, but it makes sense. Yes it is a good idea to adjust the inverter power factor. Getting the setting right will probably require getting deeper into the weeds. For example what is the power factor of the pump and when does it run? Or maybe an engineer on this forum can give you an educated guess based on the numbers you posted.
 
I think part of the problem is that if the pump is not running the energy goes through the local transformer and then is stepped down before being consumed at one of the other areas of the farm.
 
It is quite simple math.
The current and thus VA attributable to the pump motor load can be considered as two parts, one purely resistive and one purely reactive, specifically indctive,
The default inverter mode will supply only reactive current' leaving the POCO connection to supply the entire reactive portion and a smaller part of the reactive portion. This means that the worst case power factor will be far lower than that of the pump load alone.
If there is a penalty on the bill for low power factor this will trigger it. But it would not explain an absolute increase in kVAR.
Setting the inverter to deliver some percentage of reactive power will compensate for the reactive power at the pump while it is running.but will look like a capacitive reactive load when the pump is off. This may be more of a problem for POCO and the customer than the high inductive VAR, juat like those caused by overly large PFC capacitors.

This is one reason that the Hawaiian solution puts the inverter power factor under the remote control of the utility.
 
I could see the power factor going down if you have a reactive load (the motor) and a PV source not supplying the VARs needed by the motor. But I have a hard time seeing the VARs actually go up because of this.

There were some 3400 additional kVARs used in the billing period. There are only 730ish hours in a month. This suggests a new and rather large VAR load added.

Jon
 
winnie said:
I could see the power factor going down if you have a reactive load (the motor) and a PV source not supplying the VARs needed by the motor. But I have a hard time seeing the VARs actually go up because of this.

There were some 3400 additional kVARs used in the billing period. There are only 730ish hours in a month. This suggests a new and rather large VAR load added.

Jon
Click to expand...
That’s what I’m thinking also.
The inverter is set to unity?
the kVarh more than doubled?
reads like someone added something after the install.
 
GoldDigger said:
It is quite simple math.
Click to expand...
I think some of your resistives got turned into reactives in that post.

I'm curious if it's very common for a PV inverter to support monitoring at an upstream point (say the service entrance) along with dynamically adjusting power factor to keep the upstream net power factor within a target range.

Also, if a PV inverter is producing its maximum rated real power, does it typically have any capacity left to also produce some reactive current? Or would it have to curtail real power somewhat to keep the total current below a ceiling?

Cheers, Wayne
 
It's certainly possible I don't understand this correctly at all. But is it possible that the extra vars are an artifact of the metering method? For example if vars are calculated by subtracting real power from apparent power, but real power is 'negative' -- i.e. exporting to the grid -- then the calculation for vars can result in a number that's greater than apparent power. Right?
 
This is interesting. It's common for the percentage of VAR/W to go up after a PV install. The PV system only supplies W while the VAR load remains the same so the utility sees an increase in the ratio and that sometimes triggers issues with the utility. But I've never seen the VAR go up after an install. I have to wonder if the inverter set to generate a constant VAR value is providing excess VAR when the pump is not running and that is being seen by the utility as an increase in VAR load?
 
jaggedben said:
subtracting real power from apparent power, but real power is 'negative'
Click to expand...
I think you're onto something there, although the numbers don't work out nicely here. It really sounds like the OP needs to put a real power analyzer on the pump and on the service and see what's going where.

Is the KVAR bill expected to be large anyway? Sounds like some point-of-use PF correction is in order.
 
jaggedben said:
It's certainly possible I don't understand this correctly at all. But is it possible that the extra vars are an artifact of the metering method? For example if vars are calculated by subtracting real power from apparent power, but real power is 'negative' -- i.e. exporting to the grid -- then the calculation for vars can result in a number that's greater than apparent power. Right?
Click to expand...
Certainly possible...🤔

I wonder if the utility was notified of this install beforehand?
we have several PV installers connect to the grid and not inform us of the back feed.

That is, until their bill goes way up. Some double...
 
pv_n00b said:
This is interesting. It's common for the percentage of VAR/W to go up after a PV install. The PV system only supplies W while the VAR load remains the same so the utility sees an increase in the ratio and that sometimes triggers issues with the utility. But I've never seen the VAR go up after an install. I have to wonder if the inverter set to generate a constant VAR value is providing excess VAR when the pump is not running and that is being seen by the utility as an increase in VAR load?
Click to expand...
At this point the inverter is not generating VARs. It is set to power factor of 1
 
wwhitney said:
I think some of your resistives got turned into reactives in that post.

I'm curious if it's very common for a PV inverter to support monitoring at an upstream point (say the service entrance) along with dynamically adjusting power factor to keep the upstream net power factor within a target range.

Also, if a PV inverter is producing its maximum rated real power, does it typically have any capacity left to also produce some reactive current? Or would it have to curtail real power somewhat to keep the total current below a ceiling?

Cheers, Wayne
Click to expand...
Typically the input stage limits the usable real power and the output stage limits the load current regardless of PF.
So in most cases when the inverter output is limited by the solar input the reactive kVARs are free.

Sent from my Pixel 4a using Tapatalk
 
GoldDigger said:
Typically the input stage limits the usable real power and the output stage limits the load current regardless of PF.
So in most cases when the inverter output is limited by the solar input the reactive kVARs are free.
Click to expand...
That wasn't quite my question. In terms of the framework you described:

When the input stage is maxed out on real power (ample solar input, clipping), is the resulting current at nominal voltage with unity PF equal to the current limit of the output stage? Or does the output stage have further current headroom to allow non-unity PF at full rated real power at nominal voltage?

Cheers, Wayne
 
wwhitney said:
Also, if a PV inverter is producing its maximum rated real power, does it typically have any capacity left to also produce some reactive current? Or would it have to curtail real power somewhat to keep the total current below a ceiling?
Click to expand...
From the SMA data sheet for the STP62-US-41:
AC Nominal Power 62500W
Maximum Apparent Power 66000VA
 
I'm pretty sure that the inverter's maximum continuous rated output current is the hard line you can't cross if you want to get your UL listing. Therefore if the inverter is outputting its maximum current, then in order to do so at a lower power factor than unity it must necessarily limit real power to whatever the power factor setting allows. (Or at least, it can't do otherwise for long enough to count as 'continuous'.)

I don't think that 'full rated real power' is a hard line in the same way. Which means that voltage is a part of the question. For example, if the grid voltage goes up then there will be more 'headroom' for real power at a given power factor, and if the grid voltage goes down, then there will be less. That's actually true for any power factor including unity, but you get the idea.
 
Last edited:
As ggunn points out some inverters will have specs on this. But note that 'nominal power' probably isn't an engineered limit. That spec to me implies that the inverter can output up to 66000W, real power, if both the grid voltage is high enough and the power factor is set to unity.
 
I have to go but I want to throw this out there before I start thinking on this..

say the farm uses 50 kW power and 16kVar reactive.
The solar is 25 kW at unity, so now the kW has been offset but the kVar still has to be supplied by the utility. This offset is calculated by the meter as a very bad PF.
So instead of 50kW and 16kVar, your now billed at 25kW and 16kVar.
So you went from .95 to .84 pf according to the meter
 
Last edited:
Status
Not open for further replies.
Top