Transformer Load Balancing

[Wes Faul]

OK, I have another question related to the one I had yesterday. Let me
explain where I'm coming from. I'm writing an enhancement to a Utility
Billing program (for PUDs, cities, etc...) that will track transformers.
Part of that is a 'load balancing' report. The example of a 'load
balancing' report that I have gives actual usage (in KWH) of all the meters
connected to it, converted to KVA and the % it is of the maximum KVA rating
of the transformer. My question is - should I be concerned with the actual
usage or the potential usage? My guess is that if every meter reads the max
usage, it will overload 99% of their transformers. Is this typical in the
energy industry? Are the transformers usually balanced with the actual
usage or the potential usage?
Thanks,
Wes

 

[Tim Perry]

OK, I have another question related to the one I had yesterday. Let me
explain where I'm coming from. I'm writing an enhancement to a Utility
Billing program (for PUDs, cities, etc...) that will track transformers.
Part of that is a 'load balancing' report. The example of a 'load
balancing' report that I have gives actual usage (in KWH) of all the meters
connected to it, converted to KVA and the % it is of the maximum KVA rating
of the transformer.

you cannot convert kWH to kVA except as an average power used over the
period in question.

the maximum kVA rating of a transformer is of interest to the engineers and
electricians who install the service but i can not see how it would be of
any interest to a billing program.

My question is - should I be concerned with the actual

usage or the potential usage?

potential usage is 100 % of the power grid. at that point fuses open
hopefully before cable melts.

My guess is that if every meter reads the max usage, it will overload 99%

of their transformers.

watt-hour meters do not read "max usage' they read watt-hours. the meter
will have a max amperage rating suitable for the service load. (i.e. 200A,
400A 800A and so on)
some W-H meters used for industrial purposes are called "Demand meters".
these show the peak usage over the meter reading period. it is my
understanding that these are used by power companies to adjust the rate
(higher) for those users who may place a heavy load on the power grid during
peak demand hours. in others words a user with 800A service who never
exceeds 200A during the billing period gets a better rate.



Is this typical in the

energy industry? Are the transformers usually balanced with the actual
usage or the potential usage?

unless we are taking about the sub-station transformers there is nothing the
power company can do other then provide power. the balancing part is done
(or not done) by the user. a competent electrician will attempt to wire
circuits in such a way as to evenly distribute load among the branches or
phases but there is no accounting for what the user will switch on or off or
what equipment will fail.

my response here may be simplistic and incomplete, or i may not fully
understand the question and its relation to the billing problem.

i do have the feeling, based on long experience with accounting and
management types that they will have no interest in this unless it somehow
impacts the "bottom line".

 

[SQLit]

snipped

i do have the feeling, based on long experience with accounting and
management types that they will have no interest in this unless it somehow
impacts the "bottom line".

Short, sweet and to the point. For my view RIGHT ON.

Your application could vary

 

[Wes Faul]

OK, let me clarify a few things. We are just providing our customers a way
to determine if they need to rewire, install new, or change out to higher
KVA transformers. We are not billing this in any way.

you cannot convert kWH to kVA except as an average power used over the
period in question.

the maximum kVA rating of a transformer is of interest to the engineers and
electricians who install the service but i can not see how it would be of
any interest to a billing program.

The kWH is the monthly usage, and we are converting it to average kVA
(dividing the kWH by 720 - 30 days/month * 24 hrs/day). I guess the purpose
of this load balancing is to answer the question "On average, is the
transformer overloaded?".

My question is - should I be concerned with the actual

potential usage is 100 % of the power grid. at that point fuses open
hopefully before cable melts.

By potential usage, I mean all the meters attached to that transformer are
using the maximum amount it can read.

99%
of their transformers.

watt-hour meters do not read "max usage' they read watt-hours. the meter
will have a max amperage rating suitable for the service load. (i.e. 200A,
400A 800A and so on)
some W-H meters used for industrial purposes are called "Demand meters".
these show the peak usage over the meter reading period. it is my
understanding that these are used by power companies to adjust the rate
(higher) for those users who may place a heavy load on the power grid during
peak demand hours. in others words a user with 800A service who never
exceeds 200A during the billing period gets a better rate.

Our software will allow them to bill these demand meters, but I'm not sure
if that will effect the load balancing aspect of our software.

Is this typical in the

unless we are taking about the sub-station transformers there is nothing the
power company can do other then provide power. the balancing part is done
(or not done) by the user. a competent electrician will attempt to wire
circuits in such a way as to evenly distribute load among the branches or
phases but there is no accounting for what the user will switch on or off or
what equipment will fail.

my response here may be simplistic and incomplete, or i may not fully
understand the question and its relation to the billing problem.

i do have the feeling, based on long experience with accounting and
management types that they will have no interest in this unless it somehow
impacts the "bottom line".

Thanks for the info. I think I will just include both figures on the load
balancing report - what the actual average load was on the transformer, and
what the load would be in every meter was measuring the maximum amount it
can handle.
Thanks,
Wes

 

[SQLit]

OK, let me clarify a few things. We are just providing our customers a way
to determine if they need to rewire, install new, or change out to higher
KVA transformers. We are not billing this in any way.

What ever you monitor will tell you what % of capacity you have. I think
your really looking for the rate/size of change. I would guess if your
monitoring period was say every five minutes per meter, after a couple of
years you would be able to start predictions.

Do not fall into the trap of comparing June to July. You can only compare to
the same month in different years. I set up a program for a University that
did this. After the first summer I realized that I needed an constant that
was not being recorded. For my application I also recorded temperature.
Since 2/3 of the load was a/c. The temperature constant helped us
understand spikes that were not consistent over time. Your application could
be different. Just a suggestion

 

[Beachcomber]

What ever you monitor will tell you what % of capacity you have. I think
your really looking for the rate/size of change. I would guess if your
monitoring period was say every five minutes per meter, after a couple of
years you would be able to start predictions.

Do not fall into the trap of comparing June to July. You can only compare to
the same month in different years. I set up a program for a University that
did this. After the first summer I realized that I needed an constant that
was not being recorded. For my application I also recorded temperature.
Since 2/3 of the load was a/c. The temperature constant helped us
understand spikes that were not consistent over time. Your application could
be different. Just a suggestion

It sounds like you want to do a study of loading, demand, and
diversity factors. If the customer is already being metered and
billed for demand, that is a valuable source of data for your study.

There is a well known technique described in The Electric City: Energy
and the Growth of the Chicago Area, 1880-1930
by Harold L. Platt that describes how Samuel Insull did this in the
early years of the last century for a one block area of Chicago's
North Side.

 

[Wes Faul]

It sounds like you want to do a study of loading, demand, and
diversity factors. If the customer is already being metered and
billed for demand, that is a valuable source of data for your study.

There is a well known technique described in The Electric City: Energy
and the Growth of the Chicago Area, 1880-1930
by Harold L. Platt that describes how Samuel Insull did this in the
early years of the last century for a one block area of Chicago's
North Side.

Basically, this report is to notify them which transformers are being
overloaded (on average - by month).
Unfortunately, they only have demand meters at some of the locations. I'm
trying to come up with a way to convert between the kWH that is read at the
meter to some sort of kVA to compare it against the rating on the
transformer. They came up with a conversion calculation of .0035 * kWH =
KVA. They're not sure where they got it from, but they think it was from a
federal study done back in the 60's. They said this seems to work fairly
well for single phase residential meters, but don't have a good way to
convert for single phase commercial/industrial meters or for three phase
meters without a demand reading.

Does the above book have info on how to do this?
Thanks,
Wes

 

[daestrom]

compare

Basically, this report is to notify them which transformers are being
overloaded (on average - by month).
Unfortunately, they only have demand meters at some of the locations. I'm
trying to come up with a way to convert between the kWH that is read at the
meter to some sort of kVA to compare it against the rating on the
transformer. They came up with a conversion calculation of .0035 * kWH =
KVA. They're not sure where they got it from, but they think it was from a
federal study done back in the 60's. They said this seems to work fairly
well for single phase residential meters, but don't have a good way to
convert for single phase commercial/industrial meters or for three phase
meters without a demand reading.

The problem you have is that your customer can't tell you the peak to
average loading.

After all, if we're talking just two shifts, five days a week in a
manufacturing facility, your 720 hour month is more like just 320 hours,
with a very load load during the night and week-end. Some industries have
different load profiles than others (conveyor, compressor, furnaces, all
have different demands). Commercial building load can depend on the
business(s) inside. Hours of operation, elevators to take people to work in
the morning, HVAC demands, computer rooms, etc.... All sorts of variables
to be considered that you're not supplying (and even if you do, I doubt
anyone here is willing to undertake the calculations pro bono).

Such load analyses are what EE and PE get paid the 'big bucks' for. Trying
to use a 'rule of thumb' from residential usage for commercial, isn't going
to cut it.

daestrom

 

[Tim Perry]

compare

Basically, this report is to notify them which transformers are being
overloaded (on average - by month).

this is done in the field by a lineman with a clamp on ammeter. (or an
ammeter with a chart recorder)

Unfortunately, they only have demand meters at some of the locations. I'm
trying to come up with a way to convert between the kWH that is read at the
meter to some sort of kVA to compare it against the rating on the
transformer. They came up with a conversion calculation of .0035 * kWH =
KVA.

you are beating your head against a rock here. the average power used over
30 days can not be made into peak power used over a period or hours (or
minutes of one day). the "constant" will be different for each and every
type of service. it might be "safe" to lump residential loads together but
light and heavy industry will have huge differences in peak loads.


a real world example would be an AM broadcast transmitter which is daytime
operation only. sunrise to sunset it draws a steady load... lets say 25,000
watts. sunset to sunrise it is off... zero watts (relatively speaking). the
total "on" time changes every month. you long term average will show
approximately half of the true peak load.


by the time your monthly average can show an overload condition the
transformer has already caught on fire and will have been replaced.

if it a long term constant overload the voltage drop will be such that the
user will be complaining about brown outs and burnt up motors.

They're not sure where they got it from, but they think it was from a

federal study done back in the 60's.

there were a lot of hallucinogens floating around back in the 60's

They said this seems to work fairly

well for single phase residential meters, but don't have a good way to
convert for single phase commercial/industrial meters or for three phase
meters without a demand reading.

any power usage calculation that would work for single phase would work as
well for 3 phase as long as the proper formula is used.