System Losses

[Jorgy1]

The distribution network in my country is a 4-wire star system. Most of the load in the system are single phase, 230V. The system loss figure is estimated at 25% consisting of both technical and commercial. We have tried to bring down this figure but our efforts have not yielded much. I want to know if currents in the neutral conductor due to unbalance loading could contribute significantly to the losses in the network. Thank you.

 

[jghrist]

Yes, the neutral current can contribute significantly to network losses. If all the lines were 1Ø, and the neutral was the same size as the phase, half the line losses would be in the neutral. If all lines were 3Ø and the loads at every point were balanced, there would be no neutral losses. Real systems will be somewhere in between these extremes. To determine how much losses are in the neutral, you have to model the system to calculate the currents.

There are practical and economic limits to how balanced you can keep the loads. For instance, if you need to serve 3 customers on one lateral, you could achieve lower neutral losses by making the lateral 3Ø and serving one customer per phase. This would require four conductors, though, where only two would be required for a 1Ø lateral.

 

[fredpar]

I'd like to know how much percentage of the 25% are the "commercial" losses. Also what do you mean by commercial?
Thanks

 

[Jorgy1]

Thank you jghcrist and fredpar for your contributions. We define commercial losses as all the losses that are not due to technical considerations. E.g. power theft, illigal connections, limitations of billing systems etc. We estimate commercial losses to be about 10% and technical losses to be about 15%.

 

[arango]

I would like to know what kind of methods or strategies have you tried to cut down the losses, and how much do you expect to reduce them?

 

[apowerengr]

Even 15% is a very high percentage for system losses; are there measurements to support the estimate? Typically I'd expect 1-2% in secondary/service losses, 1-2% in transformer losses and 1-8% in primary system losses for a total of 3% to 12% in the worst case systems. I assume your 25% comes from total energy purchased/generated minus total energy sold. Metering errors can skew these numbers as can forgetting to take credit for certain un-metered loads - this may be part of your commercial losses?

Other than purchasing standards for transformers you don't have much control over that 1-2%, the secondary/service losses might be cut in half but never eliminated so that leaves the primary system as the area for greatest improvement.

Correct use and placement of capacitors, economic conductor sizing, phase and feeder balancing all play a part as does primary system voltage level. You may or may not have control over the last item, at least not in the short term. I don't expect you're losing mega-watts in your neutrals due to single-phase loads.

 

[jghrist]

apowerengr,

You are probably thinking of a typical North American system with small transformers and short secondaries. From the voltage, this is not a North American system. In many parts of the world, large transformers are used with very extensive 230/400 volt secondary systems. It is not unusual in less developed countries to see hundreds of customers on one transformer. Secondary system losses are higher in these extensive systems than in the typical North American system. Balancing loads on the secondary can be an inexpensive way to significantly reduce losses.

 

[waross]

What size are your distribution transformers, and how many customers do you typically have on a transformer?
Do you have long secondary cable runs?
respectfully

 

[fredpar]

Jorgy:
The reason I asked about “Commercial” losses is because I remember a utility company dealing with the same problem in South America. They came out with a very well planned campaign to stop power thief and have people paying for electricity.
There were a lot of people stealing power:
1) The most common and most traumatic was the very poor people with almost zero income. They even didn’t have a meter! In this case they changed to different residential rates. They went from the very fine expensive neighborhoods [rate#5 the highest] to the very poor ones [rate#1]. They made it possible for some really very poor people not to pay. BUT ANYWAY THEY HAD TO HAVE A METER.
2) They were a lot of businesses [stores, malls, shops] stealing power with the help of electricians and in some cases with the same power utility company electricians.
3) They did check every single meter installation in the whole state; one of these guys working in this project spent a whole afternoon telling me about his findings and I lost count of all different ways people/businesses were stealing power. They did things like putting a little bit of sand inside the meter to make it spin slower! They had what they called double wire service, and then it was possible for the larger conductor to bypass the meter and go straight to the stove. The meter was metering but a lot less.
4) Public/government agencies that didn’t pay.

Now for technical losses jghrist and apowereng suggestions are great. Can you please answer Waross and Arango’s questions: What are you guys doing so far for technical losses?
What wire sizes do you use for: primary, secondary and services to business and houses? How many customers per transformer and what sizes, also average distance to customers?
In closing I want to mention couple things: One is that in some places here in the States it’s illegal to cut the service due to no payments, they do have special consideration for the elderly and infants especially in places with extreme weather. Also let me add that most utilities here in the US have what they called the “Good Neighbor Fund” and this is to encourage people to pay additional one or two dollar a month to help to pay for the needed. It’s good and it shows the tremendous solidarity of the American people, which is pretty much the same you find in this forum.

 

[Jorgy1]

Jghrist, you are right about the nature of our distribution system. Our trafo sizes range from 50KVA in the rural areas to 500KVA in the city centres. They are usually overloaded to about 150%. We estimate that the average load per customer is 3KVA, so you can imagine the number of customers on each trafo. The LV lines (230/400V) are generally long. The longest is about 80 spans (average of 50m per span).

We agree that the technical losses are high even though we have not actually measured it. We know that the total loss is 25% from purchase and revenue figures. We have done a few things to address the technical loss problem:
1. We now test all trafos to ensure that their losses do not exceed a certain figure.
2. We are reducing lengths of LV lines by limiting trafo sizes to 200KVA. Eventually smaller trafos will be used and LV lines eliminated altogether.
3. Service conductor sizes have been increased to 16mm sq. Cu.

Some people are not convinced that unbalance loading could contribute SIGNIFICANTLY to system losses. Is there anyway to prove this considering the fact that the load is distributed? We do not usually run single phase i.e the LV lines are 3-phase and a neutral with the loads distributed along the length of the feeder.

Thanks.

 

[jghrist]

The best way to see if unbalanced loading contributes significantly to system losses is to model the system. You can use a commercial distribution analysis program or make the calculations with a spreadsheet.

You will have to estimate the load per customer. Assume a constant current load. This is probably closest to actual and makes the analysis easier. With constant current loads, a 1% drop in voltage results in a 1% decrease in kVA load. This is generally how mixed distribution systems respond. There will be a mix of constant power loads like motors and constant impedance loads like incandescent lights. You will have to start at the end of the line and add the currents vectorially at each service and tap to determine the current in the neutral.

The major cost may be a field survey to determine what phases each customer is connected to. You could model one typical secondary system, then change the model to see how much loss could be saved by connecting the customers to achieve maximum balance.

 

[waross]

I would not expect large neutral losses in the rural areas.
An estimated demand of 3 KVA at 230 volts is 13 amps.
If the customers are evenly distributed, the neutral currents will cancel after three connections. The neutral current could be expected to be a maximum of 13 amps. (13 amps added vectorially to 13 amps, 120 degrees displaced, is 13 amps.)
Compare this to the line current, make an allowance for smaller sized neutral conductors, and then decide if you want to continue with a full survey.
Look for places where the load is NOT balanced.
I would send a technician out with a clamp-on ammeter on a pole to check neutral currents in any area where uneven loading is suspected.

We do not usually run single phase i.e the LV lines are 3-phase and a neutral with the loads distributed along the length of the feeder.

As long as you can maintain this criterea, I doubt that the neutral losses will be significant.
One place that I would expect significant losses would be long, undersized service drops. You will get losses on both the hot wire and the neutral. If you can move the meters to the poles, then the customers will pay for these losses.
respectfully

 

[jghrist]

I was a consultant on two projects in Egypt that dealt extensively with low voltage networks. One was a study to determine the extent and source of losses in the distribution system. The other was rehabilitation of the electric distribution system in Alexandria. My experience is that although the phase loads can be balance fairly evenly along a low voltage line, they often aren't in practice. This unbalance represents an inexpensive way to reduce losses by reconnecting customers to increase the phase balance. The review of the system to check and correct unbalance often will result in a decrease of "commercial" losses as well because inspection of the network will find instances of theft.

 

[waross]

Hello jghrist;
Thank you for sharing your experiences. My comments were based on Jorgy1's information that the loads were distributed.
I agree with you that it would be well to check to see if the loads are in fact distributed, and if not, to balance them.
My third world experience was with a North American style system.
13,200V three phase primaries and 120/240v single phase secondaries.
Neutral losses were only a factor on the primary and were minimal.
The largest losses were theft and service drops. The utility allows #6 aluminum triplex for up to 100 amp services. Some of the runs are several hundred feet long.
respectfully