alt.energy.renewable

"Fran" < @ > wrote in message
news: @ ...
> On Jul 28, 9:58 am, "daestrom"
> wrote:

>>
>> > Which are largely offset by lower transmission losses over great
>> > distances, and the benefits of a more stable network.
>>
>> No. This is a common fallacy, that a large number of small generation
>> units
>> increases grid stability.
>>
>> To truly understand this you must define two distinct terms, 'stability'
>> and
>> 'reliability'. 'Reliability' is the chances of a major power outage
>> happening as a result of some predictable failure. 'Stability' is the
>> ability of the grid to maintain the key parameters of voltage and
>> frequency
>> to each load that is supplied during all sorts of conditions of both
>> normal,
>> abnormal and emergency operation.
>>
>> A large number of small generating units may seem more 'reliable' since
>> any
>> single generating unit failure is less likely to cause an overload and
>> collapse. That's the theory. But the details are crucial. If you swap
>> one
>> large generating unit that is more than 95% reliable with ten units that
>> are
>> only 85% reliable, have you really improved?
>
> That's the key question though. There's no reason to assume that
> smaller units will be less reliable than larger units. Depending on
> how they are maintained, it might be that smaller units are looked
> after better, or it might not.
>
>> Current technology requires
>> that the system 'survive' a loss of the largest generating unit. Since
>> generating units are somewhat centralized and large, this means there
>> must
>> be sufficient 'spinning reserve' to absorb the transient of losing such
>> large units. If there were no 'large' units and a larger number of
>> 'small'
>> units were used, then yes the transient of losing a small unit is less
>> severe and easier to compensate.
>>
>
>
> Exactly. And remember, a large unit is a much better target for attack
> from criminals than a small unit.
>

And more easily secured :-)

>> But many large-scale blackouts in the past were not caused by a loss of a
>> 'large' generating unit (at least not initially). The initiating event
>> for
>> the August 2003 blackout for example was caused by a transmission line
>> issue. If a large number of 'small' generating units are interconnected
>> by
>> transmission lines that are just as unreliable as before, then the
>> chances
>> of a major blackout are probably about the same.
>>
>
> That's true, but the effect of those transmission failures on the
> system as a whole will be smaller.
>

Not if the network is designed to operate near its optimal loading.
Overloading a 10 MW line is just as likely to cause a failure as overloading
a 800 MW line. And as Don pointed out, a 10 MW line may be more prone to
failure just because it is a low voltage line built to lower reliability
standards.

>> When it comes to 'stability' a large number of independent generating
>> units,
>> strong together over several hundred square miles with transmission
>> systems
>> is very difficult to keep stable. Power flow over a dense network of
>> inter-connecting lines means that power from each generating unit has
>> multiple pathways to reach the various loads. Controlling just how much
>> power flows over each line (to avoid overloading any given line) is
>> complicated now. Increasing the number of generating units and
>> transmission
>> pathways by a an order of magnitude will not make power flow, voltage and
>> frequency *more* stable. If anything, it will be less stable.
>>
>
> Again, it depends on a whole range of factors. Localised demand
> management is a pretty important tool, since a lot of power usage is
> potentially elastic -- one can choose when to pump water to catchments
> and when to stop. Water heaters can be adjusted to take advantage of
> times when demand is low, realtive to supply.
>

True, but outside the scope of this discussion.

>> On the point of transmission losses, they would not be lower for a
>> de-centralized scheme. By its very nature, a decentralized scheme would
>> have more transmission lines, not less.
>
> True, but each unit of energy might have to travel, on average, less
> far before being converted into work, Thus, if the typical
> transmission distance is under 50 km, losses in transmission are going
> to be less that if it's three or four times that. As I understand it,
> the need to avoid brownouts at the end of the line means having more
> power at the source than those near to the source actually need.

It is much more likely that the 'average' unit of energy would have to
travel *further*, not less.

One approach would be to group a large number of small generating units
around some central transmission station where their individual outputs
would be combined for transmission over some major line to the load (a major
city). The 'spurs' of such a spoke-wheel system would represent an
additional line whereas the traditional system would just locate the large
generating unit at the center of such a collection and transmit directly on
the major line. The 'spurs' increase losses and reduce transmission
reliability (even if net generation reliability is constant).

Or you might imagine a circular net arranged around the load itself. The
lines near the central load would be larger than the lines around the
periphery, but they would be carrying a larger load since power from the
peripheral units would combine with the power from units near the central
core as the power flowed towards the center. But think about this a bit, if
you spread the small generating units evenly over the land area, more of the
units would be near the periphery than near the center (land area rises with
distance from central load, for a circle, dA/dr = 2*pi*r). So the 'average'
distance that power must traverse is not exactly 1/2 the radius, but further
out then that. Further out means 'average' distance that a unit of energy
travels is higher, not lower.

Further distance traveled for an 'average' unit of energy means more losses
and more susceptibility to line failure ( . lower reliability).

Of course if the distribution is such that you cram the units together close
to the load and spread them out as you get further from the load, you could
mitigate this. Taken to the limit, you end up with all the generation right
on top of the load ( . a centralized plant right in downtown). Of course
there are practical limits to this and you end up with some 'acceptable'
distance from the load where you try and crowd all the generation into as
small an area as possible. Ooops, we're right back to centralized
generation.

daestrom