# Negative power factor and PV systems

Status
Not open for further replies.

#### gar

##### Senior Member
110607-1048 EDT

A standard question by A. D. Moore in his class was ---
What is the maximum phase shift in an RC network?
The usual, but incorrect, answer was 90 deg. The incorrectness can be easily illustrated by examining the fundamental design of an HP RC oscillator.

But it is true that 90 degree is correct for a single stage RL or RC circuit, meaning one R and one C or L. The trap is the word network and its meaning.

As I have stated previously in this thread the major problem is definitions.

Personally I do not think negative power factor as defined by tallgirl has much use other than to say what direction power is flowing. As a power company seeing negative PF at some point in the system what am I going to do with this information. I, as the power company, could put switched capacitors at that point that I control to modify the power factor. But would I, no.

Suppose the load is not generating any power and the PF = 0.9 . As this load starts to generate only real power, the PF seen by the utility goes down, then 0, then negative. All this time the utility was supplying the same amount of reactive power, and gradually supplying less real power, and finally receiving real power instead of supplying it.

Ideally, ignoring other minor factors, the utility would like the load to have eliminated the reactive component. Then the PF would be 1 or -1 using tallgirl's definition. And it only becomes a power direction indicator. It becomes indeterminate at 0 power flow.

I think that working with loads of this type I would rather use independent measurements of real and reactive power to evaluate the load.

I do not like PF being defined using the cos because this is only a special case of the general definition of PF. The cos special case works with linear loads and sine waves. This certainly is not the real world today. I have a much flattened peak on my voltage waveform. In turn this implies a lot of capacitor input filter loads on the system pulling large peak currents at the voltage peak. cos serves no purpose here.

.

#### Besoeker

##### Senior Member
Frustration?

The problems with PV/DRG power factor are unlike those in other situations because the impact to the grid is completely different, the money is different, voltage drop / gradients work differently, etc.
Look at post #104.

#### Smart \$

##### Esteemed Member
Look at post #104.
...

In reality, if the inverter is producing output power at unity power factor, it doesn't matter if it is going into a passive resistor or into the utility supply.
Debated matter here seems to be conventional wisdom versus neo-wisdom. The former weighs heavily on neo-wisdom making sense to all.

Your analogy uses, as convential wisdom does, the PV system as the reference. I believe the perspective of tallgirl's premise is the utility supply being the reference for PV output. From that perspective, the PV system is a load, not the source.

#### Besoeker

##### Senior Member
Your analogy uses, as convential wisdom does, the PV system as the reference. I believe the perspective of tallgirl's premise is the utility supply being the reference for PV output. From that perspective, the PV system is a load, not the source.
Which is, of course, nonsense.

#### Smart \$

##### Esteemed Member
Which is, of course, nonsense.
As was the theory the Earth is round at one point in time.

#### Besoeker

##### Senior Member
As was the theory the Earth is round at one point in time.
But we now describe it better as a geoid:

#### Smart \$

##### Esteemed Member
But we now describe it better as a geoid:

And the consensus response if you "suggested" this description a couple millennia ago?

#### Besoeker

##### Senior Member
And the consensus response if you "suggested" this description a couple millennia ago?
Dunno.
I'm not THAT old!!!

#### kingpb

##### Senior Member
and nowhere in the graph would the pf be negative!:jawdrop:

#### mivey

##### Senior Member
In the DRG world, it really does mean something completely different.
I'm sure some see it that way. You seem to.

The sign really does matter.
I suppose if it matters to you then it matters to you. I can't argue about what matters to you.

I'm sorry this is so hard for you to grasp, but I haven't the time to waste any further explaining it to you.
I understand what you are saying. I'm saying you are wrong. But by all means don't think I'm telling you you have to do it a different way. I'm just posting my opinion and trying to keep you from teaching others what I consider to be poor terminology.

For a free clue -- figure out the money aspect. What's the financial difference between importing 1KW at 90% power factor and exporting 1KW at 90% power factor.
The difference can be quite drastic.

A placeholder for "delivered" and "received"? Is "volts" now a placeholder for "how much you get shocked"?
Sigh

As an added bonus, I'm going to point out that you're now using "final power factor", not the one based on cos(Θ), which is a completely different animal.

For anyone who isn't bored, please tell them the phase angle relationship between current and voltage for a system having a final power factor of 90%. Feel free to pick "leading" or "lagging", as the mood may strike you.
Define final power factor. I'm thinking you have not provided enough information but who knows what rules you are playing by.

#### mivey

##### Senior Member
Frustration?
Please don't get frustrated. It is just a discussion. It will not decide the fate of the world.

The problems with PV/DRG power factor are unlike those in other situations because the impact to the grid is completely different, the money is different, voltage drop / gradients work differently, etc.

Yes, of course "power factor" has meanings that are well-established for many decades. And now people have to adjust to different meanings because things are different.
I appreciate that you think this is all new to the world, but I can assure you it is not.

Something like negative RMS current, which is numerically impossible, has to be considered because otherwise we wind up with "received" and "delivered" which somehow behaves exactly like "+" and "-". If I stick with a traditional definition of RMS current, I can't calculate bus bar currents because I can't put a minus sign in front of "delivered" current. And I can't even calculate KWh consumption because "power" is stuck being positive. Or "received" and "delivered", which again happens to behave just like "+" and "-". So when someone says a sign that happens to act just like a sign is a "placeholder", I'm not convinced they are paying enough attention to respond intelligently. Which is very disrespectful.
No disrespect intended. The simple fact is that power factor is just not defined that way. Also, power factor is not delivered, nor is it received. However, it is calculated using a quantity that can be delivered or received.

You want a real example? Right now, this minute, L1 Irms at the PV interconnect is 2.4 amps, L2 is 3.4 and Irms at the A/C breaker is 16.2 amps. What's Irms at the lugs from the service?
And?

I don't need to read some non-technical paper written by someone at Landis+Gyr.
Actually, it is part of L&G' s technical documentation for their meters and is included as a section in their technical documentation. Don't judge a book by its cover.

I've spent the last several weeks with my face buried in manuals from other vendors -- http://www.electroind.com/pdf/100s-manual.pdf, http://www.ccontrolsys.com/ww/images/2/28/WNC-MODBUS-Manual-V16.pdf,
And did that teach you that power factor can be a negative or that the pf values are reported with a sign to indicate a generating or motoring condition?

and if I can find a similar manual for the Landis+Gyr meters Oncor has been installing all over Texas, I'll bury my face in that.
If they are the MAXsys Elite meters, I can get you a link.

More than manuals and talk about "placeholders", I need numbers that work and make sense. Negative RMS current, as dumb and wrong as it may seem, works.
Did you read the document I linked?

Not for a final power factor value it isn't. It could be a switch-mode power supply that conducts during 90% of the cycle with no phase angle change in the current wave form.
So now we want to bring in harmonics? How many sinusoidal formulas are you going to need to use cosine now? Meters use the instantaneous data.

#### mivey

##### Senior Member
But we now describe it better as a geoid:

Thanks. That was actually some interesting reading.

And the consensus response if you "suggested" this description a couple millennia ago?
Off with that heretics head!

Last edited:

#### Helios

##### Member
powerfactor +/-

powerfactor +/-

A solar PV inverter can provide current in phase with a grid or either positive or negative (leading or lagging) (see: http://en.wikipedia.org/wiki/AC_power#Power_factor)
Inverters can work to adjust the power factor day and night by the way, and utilities are looking into using inverters to help adjust the power factor - not today. It would only be practical for larger inverters (10-100MW), it is easy to do, just a software change in the inverter.

#### mivey

##### Senior Member
Let me try to make a different illustration. It has been a while since pressure vessels, etc and I did not feel like looking up the specifics so just bear with the analogy even if there are some specifics to flow and pressure vessels that may be a bit off.

Let water flow through a pipe mainly in one direction. Due to some pipe flaws (inefficiencies) we also have some oscillatory eddy flow that moves in both directions. We have flexible section of pipe that will maintain a constant pressure. We need to know how much to allow for that joint to swell so we can allow for clearance.

With eddy flow added to the uni-directional flow, the clearance diameter must increase to what we can call the normal clearance diameter. If we can get rid of the eddy flow, we can minimize the clearance diameter that we need for a given uni-directional flow (i.e., fewer water molecules in a given cross sectional area of pipe).

This minimum clearance diameter divided by the normal clearance diameter is an indicator of how efficiently we are utilizing the space needed. This ratio (or percentage) is like the power factor in the electrical world.

Suppose that we reverse the main flow of water (like we might do in the electrical world with a generator pushing power onto the grid). Again, in the normal flow case, we have a large pipe clearance diameter because we are handling the uni-directional flow as well as the eddy flow.

Again, if we can get rid of the eddy flow, we can minimize the clearance diameter. This minimum clearance diameter divided by the normal clearance diameter is a measure of how efficiently we are utilizing the space needed. The percent diameter has nothing to do with the direction of the uni-directional flow. It is just an un-signed percentage.

Now to provide more information about what is going on with the rest of our setup, we can assign a "+" or "-" symbol that associates the percentage with a particular direction of uni-directional flow, but that does not mean the percentage itself is a negative or a positive. I believe this is what Fluke does with their PQ meters (they use a capacitor and inductor symbol to indicate leading and lagging and the +/- to indicate the direction of real power flow). I think that is what tallgirl is talking about.

We can also assign a "+" or "-" to the percentage to indicate if the eddy current is in the same or opposite direction as the uni-directional flow. That would be similar to the use of an indicator for leading or lagging (some meters use +/- to indicate leading and lagging).

Either way the symbol is used, it still does not mean the percentage itself is a negative or a positive.

The following graphs show a real, reactive, and real+reactive load. In the final case, the generator supplies the kW load plus pushes an equal amount of kW back on the system. This is the opposite real power case of graph #3. As you can see, the ratio of real kW to kVA is the same and thus the efficiency of #3 and #4 (thus the power factor) is the same. The kW direction did not matter.

Real Power Load graph:

Real + Reactive Load:

Reactive from utility + Real to utility(generator supplies the real portion of the load plus pushes an equal portion of real power back on the grid:

#### tallgirl

##### Senior Member
No disrespect intended. The simple fact is that power factor is just not defined that way. Also, power factor is not delivered, nor is it received. However, it is calculated using a quantity that can be delivered or received.

The problem is that "delivered" and "received" aren't signs. They don't work with math, and they don't reflect revenue or expense.

Just as RMS Amps (volts and watts ...) are always positive values, the reality is that "RMS Amps Delivered" behaves differently than "RMS Amps Received". They behave so differently that if one looks at the numerical values in a nodal analysis, "delivered" and "received" happen to have opposite signs.

You just plain can't do math on a "delivered" or a "received" and if you're incapable of comprehending that "5 amps delivered" and "5 amps received" are two very different critters, there's nothing at all I can do to help you.

ANYWAY, I've been busy. Sorry I've not been around to 'splain cipherin' to you.

#### Electric-Light

##### Senior Member
(some meters use +/- to indicate leading and lagging).
Why do you assume that inverter provides perfectly symmetric sinusoidal output?

You could feed half wave power into the grid using diodes, have zero phase shift and still have very low PF.

#### mivey

##### Senior Member
The problem is that "delivered" and "received" aren't signs. They don't work with math, and they don't reflect revenue or expense.
Neither does the percent power factor by itself.

Somehow, as ignorant as some meter manufacturers must be because they do not have a sign on the power factor value, the ignorant POCO folks still managed to bill the customer correctly. I assure you it was not just dumb luck.

Just as RMS Amps (volts and watts ...) are always positive values, the reality is that "RMS Amps Delivered" behaves differently than "RMS Amps Received". They behave so differently that if one looks at the numerical values in a nodal analysis, "delivered" and "received" happen to have opposite signs.
And the choice of sign/direction is completely arbitrary.

If you want to look at it another way, the power factor is a measure of the percent of the used conductor path that is transporting real power. The percentage does not depend on the direction of power flow. In other words, +8 kW divided by 10 kVA means 80% of the path is used to transport real power. -8 kW divided by 10 kVA means 80% of the path is used to transport real power. Either way, it is 80%. The reverse flow does not mean we use -80% of the conductor path.

You just plain can't do math on a "delivered" or a "received" and if you're incapable of comprehending that "5 amps delivered" and "5 amps received" are two very different critters, there's nothing at all I can do to help you.

...The percent ... has nothing to do with the direction ... It is just an un-signed percentage.

Now to provide more information about what is going on with the rest of our setup, we can assign a "+" or "-" symbol that associates the percentage with a particular direction ... but that does not mean the percentage itself is a negative or a positive.

... I believe this is what Fluke does with their PQ meters (they use a capacitor and inductor symbol to indicate leading and lagging and the +/- to indicate the direction of real power flow).

... some meters use +/- to indicate leading and lagging

...

Either way the symbol is used, it still does not mean the percentage itself is a negative or a positive.

ANYWAY, I've been busy. Sorry I've not been around to 'splain cipherin' to you.
It is good to be busy in today's economy.

#### mivey

##### Senior Member
Why do you assume that inverter provides perfectly symmetric sinusoidal output?
Did I say that?
You could feed half wave power into the grid using diodes, have zero phase shift and still have very low PF.
OK. But does that change the fact that some meters use +/- as indicators or capacitor/inductor symbols as indicators?

Status
Not open for further replies.