Losses in HV Transmission: How Far Can Power (Reasonably) Get Sent?

[racookpe1978]

I've been in the power production (mechanical, maintenance and piping though) field for a while, but have not seen this issue addressed from a practical standpoint.

On a political web forum, we were discussing the problems with "alternate energy" and I was challenged to validate my statement that - while voltage can be "sent" over extremely long distances through the regional grids, "power" can't be transmitted for more than 800 to 1100 miles before resistance losses become too high. Thus, while some additional solar power or wind power could be produced (for limited amounts of every day) in AZ or north TX, the power produced could not be transmitted to the users further than about the Missouri or Missi rivers.

I know losses are mainly I^2R, and as voltage increases, current (thus losses) reduce by the square. Easy. But what are actual losses for today's real-world (not AGW-propagandized next world order) transmissions system?

My analogy to these non-engineers was a garden hose: You can string two 50 foot sections together and water your front yard with little problem. Spray the water out at the end, and you could put out a trash fire. String 8 sections together and you can get water to your neighbor's kitchen, but not with a great flow. You'd still have full pressure (voltage) - but only if there was no flow (power = IV). Go further down the street with 12 or 18 sections, and you can't even fill a sink or toilet, much less fill a bathtub or put out his house fire because transmission losses are so high.

They didn't buy it: Claimed that the national grid sends power "completely cross-country all the time" and "all" we need is national money for GE's "smart grid."

On-line sources show links between the regional grids only at a few spots (most apparently with HV DC lines), and that only a few dedicated lines - such as between Hoover Dam and LA/South Ca - are true long distance carriers.

What's the effective limit (or most efficient limit) for transmitting HV power today? Do we expect anything coming up that really make a difference in the next twenty years?

 

[waross]

The distance that power may be transmitted at a given voltage is often limited by voltage drop and the capability of the On Load Tap Changer at the receiving end transformer.
If the OLTC is capable of correcting for a 10% voltage drop, then the maximum current may be limited to a value that will cause a 10% voltage drop in the transmission line.
Note that the reactive voltage drop due to the inductance of the transmission line may be greater than the resistive voltage drop.

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

 

[ScottyUK]

Go for HVDC, no OLTC to worry about. Death by acronyms...

A couple of years ago ABB were working on a 2000km+ link with about 6.5GW capacity in China. That is one of the bigger ones, in terms of both capacity and distance.

This ABB document is a couple of years old but well worth a read.



----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[jghrist]

As noted by others, limits on HVAC transmission are usually based on reactance, not resistance. There is also a stability limit that is based on reactance of the line.

I do think that we should be concerned with resistive losses, however. Although they may not limit the line loading, they have an economic impact and should be counted as a cost of the alternative energy source.

A typical surge impedance loading (SIL) limit for a 795 kcmil ACSR 230 kV line is about 140 MW. Losses at this loading would be 4.84 MW for a 100 mile line, or 3.5%.

 

[racookpe1978]

3.5% on 100 mile line is much more than I expected: That would indicate I'd lose 35% of my generated power after only 1000 transmitted miles.

But, if the "grid" is as load-limited as stories indicate, I'd face the "minor" problem of building that transmission line from scratch on existing easements: that is, if the government built a "new" 2000 megawatt solar generating station in south AZ to satisfy some green power legistlation, they'd have to buy a completely new parallel set of transmission lines east from that plant towards the final users. The existing HV grid lines (in general) could not carry the extra current. Obviously, there are places where the grid is overbuilt and the plants and customers that connect to it are lagging, but I understand that case is the exception, not the rule.

Are the net inductive and resistive losses the same if that theorectical 2000 MegW solar plant in AZ provided power to west NM, the current plants in west NM provided power to east NM, the plants in east NM provided power to west TX, etc. Seems like that would be the case: the electrons don't care where they get generated, and shifting power across country from area to area creates as much loss as sending it across in its own power line.

 

[Overvoltage]

But, you wouldn't use a 230kV line to go 1000 miles. For a long haul, high capacity line like you are suggesting, you would use 500kV, 765kV, or HVDC. Any of those would reduce your losses significantly as compared to your estimate. Of course, this would make it significantly harder to "tap" into, so if the line was 500kV or above, it most likely would be "point-to-point" over that 1000 miles with no intermediate substations.

 

[jghrist]

if the government built a "new" 2000 megawatt solar generating station in south AZ

That's a bit large. According to a recent T&D Magazine article,

http://tdworld.com/substations/construction/power_ses_construct_massive/index.html,

a 500 MW plant in California (future expansion to 850 MW) will be capable of producing more electricity than all other US solar projects combined.

 

[racookpe1978]

But, you wouldn't use a 230kV line to go 1000 miles. For a long haul, high capacity line like you are suggesting, you would use 500kV, 765kV, or HVDC. Any of those would reduce your losses significantly as compared to your estimate. Of course, this would make it significantly harder to "tap" into, so if the line was 500kV or above, it most likely would be "point-to-point" over that 1000 miles with no intermediate substations."

--

Which relates to my question: Since that theorectical 500 kV (or higher) direct-service lone distance power line can't use the same wires - but could it even use the same easement? - as the current 230 kV cross country lines, you'd have to add in the extra transmission costs to any theorectical solar (or wind) "free energy" green powered plant. Add also the extra transformer costs.

===

Yes, that theorectical 2000 MW solar plant is much larger than any now planned. But it shows the problem that these people I'm talking to don't understand yet: We would be sponsoring all the costs of a 2000 MW plant in the desert just to get 1000 MW out of the wires for 6 hours a day - and still not have the power where the electricity would actually get used. And no source for the remaining 18 hours of the day.

 

[slavag]

Posted by QBplanner:

In the text book you can often see "sub-transmission or UHV,EHV .."
In the company I work (Canada) we simply call what I mentioned above. I am not familiar with system below 12kV. So can not offer any helps.

And it is hard to define the system based on distance or transferred power.

We have a 25kV system supplying loads (5MW) 100km away from sub.

But normally,

25kV is better less than 30km (Distribution, small X/R ratio)
110kV is better less than 100km (Persection).SIL is 50MW
230kV is better less than 200km (Per section). SIL is 140-160MW.
500kV is better less than 300km (per section). SIL is 980-1200MW.

Beyond 300km you may consider DC or compensated AC system with heavy transfer.

Above are just "sound" systems. In reality, we once developed 500km system in 138kv Level. We got full of problems.

THere is a classic paper written by Mr. St. Clair. talking about transfer capability via different transmission system. Becareful of the assumptions he made.

 

[itsmoked]

SIL? Is this the standard "capacity term" for transmission lines? Can someone please flesh this out a bit?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[davidbeach]

Surge Impedance Loading.

http://www.o-t-s.com/sil.htm

http://en.wikipedia.org/wiki/Characteristic_impedance

 

[itsmoked]

Excellent.
Thanks David.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jghrist]

Another thing to consider on the losses is that I calculated losses at the line capacity. Losses are proportional to the square of the current, so the average losses will be a lower percent of average load than the peak losses are as a percent of peak load. It depends on the load factor and the shape of the load curve. With a load factor of 50% (average load 50% of peak), the average loss from my previous example would be approximately (0.3·0.5 + 0.7·0.5²)·4.84MW = 1.57 MW, or 2.2% of the average 70 MW load, if the line was loaded to its limit at peak.

 

[racookpe1978]

But David: Quoting the SIL article referenced:
"Note in this formula that the SIL is dependent only on the kV the line is energized at and the line's surge impedance. The line length is not a factor in the SIL or surge impedance calculations. Therefore the SIL is not a measure of a transmission line's power transfer capability as it does not take into account the line's length nor does it consider the strength of the local power system. "

So, then is line length (resistance) "added" to the SIL (reactance) factor to get total line losses?

 

[alehman]

There is not a simple answer to this question, because the grid is just that. A single isolated line can theortically supply current up to it's thermal limit.

Discounting thermal limitations, economics and energy losses, the other major issue is the stability of the systems the line connects. Interconnected AC power systems must maintain synchronization at all times. As more load is added or the line gets longer, the ability to maintain synchronous operation becomes more difficult. If the system operated in a completely steady-state condition, theoretically you could operate the line up to the edge where synchronization is lost. Unfortunately power systems are subject to transient events such as short circuits that must be handled without loss of synchronization. The stability of a system is almost always limited by the reactance of the lines, not resistance. And for longer lines, system stability is frequently the overriding constraint.

The other stability criteria is for voltage. This gets back to waross's reply. If too much power is transmitted, voltage control becomes difficult or impossible. Loss of voltage control was a factor in the 2003 blackout in the northeastern U.S.

There are several good books on power system stability.

http://books.google.com/books?id=8J...fkrcsO&sa=X&oi=book_result&ct=result&resnum=3

Alan
----
"It’s always fun to do the impossible." - Walt Disney

 

[wfowfo]

I know this might be largely off topic, but I heard someone once discuss the "dangers" of extrememly long transmission lines when you began approaching the 1/4 wave length of the 60 hz frequency that would make the line essentially a 60 hz radio antenna.
I don't have the expertise to comment.