Costs of Reactive Power 1

[NewGrid]

Value of Lost Load (VOLL) is the value of the demand curtailed and is approximately 30,000 £/MWh. Ie this is the cost of not
supplying 1 MWh to the consumers.

However, demand is not only MWh, but it is also MVAr. Do you have an idea of the cost of not supplying 1 MVAr to the load?

Furthermore,
Is the cost of producing 1 MWh from a generating unit, equal to the cost of producing 1 MVAr from that unit?

Thank you very much

 

[magoo2]

Value of Lost Load (VOLL) is the value of the demand curtailed and is approximately 30,000 £/MWh. Is this is the cost of not supplying 1 MWh to the consumers.

I would look at this incrementally – before and after.
Before – have a higher demand. That means MW to me.
After – have a lower demand.
Can evaluate the difference. It’s easiest to me if you draw the two on the same graph and shade the area that is displaced by the demand reduction. This area will be in MWh and can be evaluated by the incremental energy cost.

However, demand is not only MWh, but it is also MVAr. Do you have an idea of the cost of not supplying 1 MVAr to the load?

I wouldn’t call this demand reduction but suppose you have a Mvar reduction. I would evaluate this against the cost of switched capacitors to suppy it.

Furthermore, Is the cost of producing 1 MWh from a generating unit, equal to the cost of producing 1 MVAr from that unit?

Supplying reactive closer to the load is usually less expensive than supplying it from generation. The resulting losses will be less also.

 

[cranky108]

If supplying capacitive MVar has a price, then a load reduction where to happen, would there also be a price on reactive MVar that has to be supplied.

 

[magoo2]

I was guessing if you had a load reduction, you would also have a reactive load reduction so you would have less Mvar to supply.

 

[ScottyUK]

Reactive power cost is principally in capital tied up in plant which needs to be sized to provide the reactive demand as well as the active power, and arguably in lost revenue from sale of active power which wasn't possible because plant was utilised meeting reactive demand. The running cost of delivering reactive power from a synchronous machine is small: if the generator is lagging then the field will be over-excited and will consume more power than if the machine were at unity PF, but the increase in field power is small fraction of overall machine power input with a consequently small cost associated with it.

Most loads don't readily allow separation of the reactive and active components to the point where you can choose not to deliver reactive power while supplying reactive power. For example, if you don't supply reactive power to an induction machine then it won't run. The question is slightly flawed, or at least I'm not reading it the way you thought you wrote it.

 

[bacon4life]

The billing cost of MVar*h varies widely, so evaluating the cost of MVar*h not served would range from almost zero to some other very small fraction of the MWh not served. In my region instead of charging for MVar*h, customers must meet a specific power factor based on their peak demand. If they do not meet the pf, their demand charge is ratcheted for 12 months.

Since the cost for real power not served is a highly speculative number, I don't see how including the much lower costs for reactive power would be useful from a grid perspective. Perhaps it would be useful if you are looking at the impact of specific capacitor bank?

 

[cranky108]

For those few of us who have a large amount of UG cables, a loss in load may also require a reactor be used to keep the transmission voltage from rising to high. Thus a loss of load may require the use of a device which has an expence to purchase and install. So a cost of MVars could be incured with the loss of load.

Granted it would need to be a large load.

 

[jghrist]

The value of load lost is not directly related to the cost of generation, or to the cost of the utility serving the load. It is related to the lost production of the facility being served. You indicate a cost of demand, but show a cost per MWh, which is energy. The VoLL is normally a cost per MW and varies depending on the length of the outage.

Estimating the VoLL in terms of Mvar doesn't make sense. You are estimating a cost that is not really a cost of electricity, but a cost of production loss. For example, if a plant loses power, employees still get paid, but there is no product produced to sell, so money is lost. In some processes, material is ruined if the power goes out. Addition costs are incurred to restart production. The production loss is usually higher for large facilities with higher load, so it can be related to load demand.

 

[waross]

Billing typically includes penalties for excess reactive power use. Look at the loss in billing revenue on a customer by customer basis.
The reduction in fuel costs for generation will include the losses associated with reactive power delivery.
Reactive power is often not billed but penalized. Look at the loss of revenue from penalty charges.

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

 

[m3ntosan]

@NewGrid,

Looks like you're talking about the UK, to get an idea, National Grid is paying power stations £3 per MVArh. In some other parts of ENTSO-E you don't have to pay for MVArs as it's part of the grid code connection agreement for power station to provide it for free, unless you're restricting their MW output by asking for to much...
In the UK you can end up in silly situations with a generator with GSU on fix tap (there was a discussion in here about how they tap GSUTx instead of adjusting the excitation) doing for instance 100 MVAr leading at the commercial boundary in a zone with poor voltage profile, basically killing the volts in the area and although is highly inconvenient you still have to pay for it!

May you grow up to be righteous, may you grow up to be true...

 

[m3ntosan]

@cranky108

If you're talking about 275 or 400 kV cables, SIL is so high that it makes almost no difference if the cable is lightly or highly loaded, still generates roughly the same amount of MVArs.

May you grow up to be righteous, may you grow up to be true...

 

[waross]

That's the point, m3ntosan. An amount of MVARs that is acceptable when the MWs are high may be unacceptable when the MWs are low. It has to do with the ratio of MVARs to MW, not the quantity of MVARs.
And don't forget that VARs are imaginary. They are an imaginary quantity that very well describes and quantifies the affect of a displacement of the current waveform relative to the voltage waveform under different circuit conditions.
Before considering the VOLL of VARs, you must consider the revenue from the VARs.
The cost of generating VARs is very low and is mostly the cost of the additional I²R losses that the increased current causes.
The cost of the equipment to produce the VARs is a fairly large one time capital cost for the purchase and installation of the extra capacity of the generating, transforming and transmission equipment.
You should consult an accountant for advice as to how to calculate and report capital costs on a monthly, daily or hourly basis.

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