T&D Conductor Sizing

[jeebusmn]

I am trying to determine if it is worth while to install a larger conductor based on savings due to reduced losses. I thought this might be worthwhile looking into because the cost difference between a 336.4 and a 1590 transmission line per a mile is only 70k.

At the company I work for, the way that this has been determined is by looking at the Summer Peak MVA flow across a section that is going to be reconductored and multipling that power flow by the load factor, 0.49. This then gives you your I^2R losses.

The powerflow program that I am using, PSSE, can give me the losses in our sytem before and after I put a transmission line in place. I put a new line in our summer peak models and it spits out what the system losses are. This part is great and I would really like to use it to determine the true savings of a larger conductor.

The problem I am running into is trying to determine what the average system power flow values are. On the spreadsheets that used to be put together, it was just assumed that the average powerflow for the year would be the load factor (0.49) times the summer peak load. I was hoping that I would be able to just reduce our generation and loads by to 49% in our summer peak model to get the yearly average losses. I tried that and the losses in our system dropped from 65 kw to 35 kw, which is less than what I expected. I suspect power from our neighbors is flowing into our system. I was expecting due to I^2*R that if I halved the loads that the losses would drop to something close to 25%. Am I missing something? I would really like to use the simulator to give a better estimate in how much energy is saved with putting a new conductor in but with the above problems I don’t know how well I can trust the result.

How do other utilities go about this? Are simulators used? Do they just put spreadsheets together? Are there simple rules of thumb?

 

[tlrols]

Many utilities use a "loss factor" which is a coefficient less than one. This loss factor is generally around 0.25 and is simply multiplied by the peak losses...the resultant value is then presumed to be what the true loss savings for the year would be. Long practice has shown this 0.25 (or so) factor to be surprisingly accurate. I wonder if the 0.49 figure you use is perhaps too high, or perhaps not...it really depends on the utility.

The process you describe for changing the loading is somewhat flawed in that you need to reduce the load in the entire case to a new loading point. If you have a heavy case and a light case you could then figure how many hours you spend in each loading region and calculate your losses accordingly. You, of course, should have 8 cases, two for each season (i.e. light/heavy load for spring, summer, fall, and winter).

Simulation is a fine way to capture loss savings and in fact is better than any other method. You should look at some SCADA information and figure out what the loading profile looks like for this line over an 8760 hour long year.

It is good that noticed the differential cost in conductors is not very significant. Frankly, any time spent buying market based energy in a perturbed market renders the transmission cost pretty meaningless. Put another way, only morons worry about transmission costs when energy wholesales for $100 MW-hr going long. Sadly there are still too many utility types who think least cost planning is based on 2 cents per kw-hr retails rates--those days are gone and they aren't coming back. No one ever complained about a wire that was too big!

 

[jeebusmn]

Thanks for replying trol.


One part that got that got left out of my post was that with the spreadsheet method a certain MVA flow is assumed. The losses are calculated based on the assumed MVA flow. I was wanting to use a simulator to calculate this because I felt this completely neglected how a conductor's size can influence how power flows through a system.


What I am calling a 'load factor' I believe is different from your 'loss factor'. If I take my summer peak load and multiply it by the load factor (0.49) I am told that this will give me the average load across the system.


After reading your post, I looked up loss factor. I came across equations that are sometimes used by utilities to estimate what their loss factor is based on their load factor. Depending on which equation you use, I am seeing a loss factor between .25 and .3 based on a load factor of 0.49.

 

[jghrist]

A rule of thumb for transmission line losses is:

Loss Factor = 0.3·Load Factor + 0.7·(Load Factor)²

See

http://books.google.com/books?id=Wu...onepage&q="loss factor" "load factor"&f=false

 

[bacon4life]

Using historical SCADA data would be the best plan, although you do have to decide whether to use the peak load from SCADA or your model. SCADA could contain high loads from short term switching for contingencies.

I just went through some loss analysis for a large 230:115 kV transformer and was surprised to find that although we have two substation ~10 miles apart with similar transformers, the loss factor on one was 0.36 and on the other was 0.23.

Is this line just serving typical loads, or is there generation on it? Any type of generation could greatly influence the light load vs heavy load split.