Line Loss Compensation for Metering

[LJP1]

I have a wind farm generation & tie-meter design where the meter location is at a wind farm and the point-of delivery for LBA and PPA settlements is several miles down a radial 345kV transmission line (at bus of transco). A few loss compensation questions have come up. Line loss only, no transformer losses need be considered as part of this installation.

1. Does the reactive impedance of the line have any impact on the calculated line losses that will be subtracted from the generation for a loss-compensated energy value (real watts and w-hr)? The argument was made that the settlements are based on 'real' power and only subject to the resistive component of the line impedance. e.g If the phase currents are measured, the per phase real power line losses really should be just I-sq-R (not dependent on X) where I is the magnitude of the phase current, for a line modeled as a series resistance and reactance with no shunt components. And per-phase reactive power line losses should be I-sq-X (not dependent on R). With that said, depending on the line construction and line charging characteristic at different generation levels, does it impact the real losses?

2. If simply a I-sq-R calculation, is it acceptable to average the meter phase currents, such that 3 times Iavg-sq-R represents the total losses. Or, should a per phase loss be calculated and then summed to get a total loss? The later would assume balanced 3-phase generation, which I think would be the case.

3. If this loss calculation was performed outside the meter, at what cycle rate should the loss calc (compensated watts) be done as to ensure an accurate integrated w-hr results (calculated accumulator and/or periodic load profile register data). For example, consider two (2) separate generators sharing the same common line to the point of delivery. This would first require reading uncompensated data from two (2) meters, one (1) for each generator. Then, losses would need to be proportionately calculated and assigned based on the real-time production of each generator. Then compensated data written back into 'compensated' analog and accumulated registers of each meter for separate SCADA and energy accounting processes.

 

[HamburgerHelper]

1. Yes. You can't move watts or vars without increase the current. Current creates losses. I^2*R = Real losses. I^2*X = Reactive losses. The greater the reactive impedance, the more vars need to flow into the line to create the magnetic fields. The greater the real impedance, the more heat that will be created on the lines due to the current. How about using a PI model to model the transmission line? A long line model would be the most accurate since it distributes the capacitance along the line but I don't know how accurate you are wanting to be for metering?

2. I doubt it will matter. I bet you have type 4 turbines, ac-dc-ac, which should be near balanced. Per phase would be the most accurate.



------------------------------------------------------------------------------------------
If you can't explain it to a six year old, you don't understand it yourself.

 

[waross]

The line Amps will be the combined reactive current and real current.
The line losses will be I[sup]2[/sup]R. Any reactive component or reactive component will affect the line current so if you are measuring line current, you can forget about real and reactive currents. The actual current is what creates heat and losses regardless of the phase angle or reactive component.
The losses will probably be equal across the three lines but good practice is to measure each phase and take the sum. This covers you in the event that an equipment failure leaves one line producing less or no output for a time.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[JJ_Roy]

Couple of questions:
1. Who owns (or will own) the radial 345kV line between the wind farm and the delivery point (POD)?
2. What does the interconnection agreement (or energy regulations) say regarding the calculation of line losses?

Not sure what part of the world you're in but I've seen a couple of cases where a HV meter set at the POD would have paid for itself two times from all the money spent on litigation, expert witnesses, back payments, etc. for exactly the same issue you describe above.

 

[bacon4life]

Trying to calculate real-time line losses outside the meter and then assign losses to two separate entities seems like a potential nightmare. Simply verifying that loss compensation inside of a meter is correctly applied still ends up being quite complicated.

For power sales across the grid, losses are normally compensated at the annual average loss rate rather than trying to assign specific losses for specific hours. Assuming both generators have relatively similar generation profiles, this approach would drastically simplify the metering. Even if the generation profiles are different, I think the simplicity is likely worth any decreased precision.

Depending where you are located, review the requirements for generators to provide control voltage services or to provide reactive power. The reactive power series and shunt losses in the transmission line may need to be included so that the generators provide the correct voltage control service.

 

[waross]

Thinking outside the box:
Losses may also be calculated by EIxPF where E is the conductor voltage drop and PF is the power factor of the line.
If the output of a CT is dropped across a circuit with approximately the same X:R ratio as the line and then both the current and from the CT and the voltage from the burden circuit are fed to the respective inputs of a KWHr meter that should give a fairly accurate indication of the line losses.
A scaling factor or multiplier will have to be determined.
I would be looking for a KWHr meter able to work accurately with low and varying input voltage.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

That seems so much more involved than simply telling the meter the system (line and/or transformer) impedance between the measuring point and the metering point. A decent meter can take it from there.

 

[waross]

You know me David. Old school all the way. grin

Bill
--------------------
"Why not the best?"
Jimmy Carter