Relationship between Maximum demand and size of a diesel generator

[T.M.Haja Sahib]

mivey:
I think the following two points emerge from the discussion so far.

1) The relationship between MD and gen set size is site specific.

2) From such a relationship, the minimum size of the gen set can be worked out.

 

[mivey]

The instantaneous is accounted for, it just does not show up in the significant digits.

It is simply because the size of the starting loads is small or the frequency of the same in the demand interval is not high.

True enough. But it would be unusual to have a bunch of re-starts in the same demand interval. If you had a motor starting every few seconds during a 15-minute interval, the metered demand would be close to the starting kW plus the normal load. Just not a likely scenario.

So the following is not always correct.

Starting together or separately (assuming the same starting kW) will give the same demand either way if all is in the same window.

Yes, it is correct. It is just simple math:

(S1 + S2 + M1 + M2 + L) / T = ( [S1 + S2] + M1 + M2 + L ) / T

where S is the individual motor start energy, M is the individual motor run energy, L is the energy for other loads, and T is the demand interval time. The energy used during the interval does not change because of the order of summation. As long as the energy was used during the period it does not change the demand. I have not looked up the thermal demand response to see if that is true for thermal demand emulation.

 

[mivey]

mivey:
I think the following two points emerge from the discussion so far.

1) The relationship between MD and gen set size is site specific.

yes.

2) From such a relationship, the minimum size of the gen set can be worked out.

Yes. Either by using the manufacturer calculation method or by obtaining the MD from something with much better resolution that the revenue meter (like a PQ meter). However, if you are concerned about warranty or performance issues, I would recommend having the manufacturer recommended calcs in addition to anything else you might do.

 

[T.M.Haja Sahib]

2) From such a relationship, the minimum size of the gen set can be worked out.

Yes. Either by using the manufacturer calculation method or by obtaining the MD from something with much better resolution that the revenue meter (like a PQ meter). However, if you are concerned about warranty or performance issues, I would recommend having the manufacturer recommended calcs in addition to anything else you might do.

I do not think the gen set manufacturers have calculations, without reference to the MD of an installation, for minimum size of a gen set suitable for that installation. Do they?

 

[mivey]

I do not think the gen set manufacturers have calculations, without reference to the MD of an installation, for minimum size of a gen set suitable for that installation. Do they?

Depends if you want to carry full load or not (one of the "other limiting factors").

http://www.cumminspower.com/www/literature/technicalpapers/PT-7007-SizingGensets-en.pdf

http://cumminspower.com/en/technical/systemdesign/

 

[T.M.Haja Sahib]

True enough. But it would be unusual to have a bunch of re-starts in the same demand interval. If you had a motor starting every few seconds during a 15-minute interval, the metered demand would be close to the starting kW plus the normal load. Just not a likely scenario.

It is unusual because you consider re-starts of only one motor. Consider starts of many large motors in the same maximum demand interval such as in a steel rolling mill. Staggering starts of such loads over other demand periods would reduce the MD noticeably.

Yes, it is correct. It is just simple math:

(S1 + S2 + M1 + M2 + L) / T = ( [S1 + S2] + M1 + M2 + L ) / T

where S is the individual motor start energy, M is the individual motor run energy, L is the energy for other loads, and T is the demand interval time. The energy used during the interval does not change because of the order of summation. As long as the energy was used during the period it does not change the demand. I have not looked up the thermal demand response to see if that is true for thermal demand emulation.

No. It is not correct. Just shift S1 or S2 or both to non- maximum demand period or see to it that the two loads do not start in the same 15 minutes period, because S1+S2> S1 or S2. Simple arithmetic!

 

[mivey]

It is unusual because you consider re-starts of only one motor. Consider starts of many large motors in the same maximum demand interval such as in a steel rolling mill. Staggering starts of such loads over other demand periods would reduce the MD noticeably.

The OP was talking about staggering loads within the same demand period. Read the posts again.

No. It is not correct. Just shift S1 or S2 or both to non- maximum demand period or see to it that the two loads do not start in the same 15 minutes period, because S1+S2> S1 or S2. Simple arithmetic!

Completely different discussion. Re-read the posts and you will see the discussion was about one demand period.

 

[T.M.Haja Sahib]

The OP was talking about staggering loads within the same demand period. Read the posts again.

Completely different discussion. Re-read the posts and you will see the discussion was about one demand period.

You are correct. But still there is something wrong.

During a maximum demand period, the starting together or separately of various loads does not matter. But starting together of various loads in an initially non-maximum demand period may make it a maximum demand period and maximum demand meter may then record an increase in its reading.

 

[mivey]

You are correct. But still there is something wrong.

Nothing is wrong, that's how demand metering works.

During a maximum demand period, the starting together or separately of various loads does not matter. But starting together of various loads in an initially non-maximum demand period may make it a maximum demand period and maximum demand meter may then record an increase in its reading.

So, your revelation is that a demand meter is going to record the interval with the highest average demand. Well how about that!

 

[T.M.Haja Sahib]

a demand meter is going to record the interval with the highest average demand.

My point is that, even when the average power of a starting load in a demand interval ( please note that not its average over the demand interval is meant here) is higher than previously recorded MD, that demand interval will become a maximum demand interval for the MD meter.

 

[gadfly56]

My point is that, even when the average power of a starting load in a demand interval ( please note that not its average over the demand interval is meant here) is higher than previously recorded MD, that demand interval will become a maximum demand interval for the MD meter.

It is, of course, theoretically possible that shifting a load from one demand interval to another will change the current MD window. But you would already have to be running so close to the current MD that the tiny fraction would push you over. In mivey's calculation, the inrush kW was about 0.04% of the total kW over the demand period. I wouldn't worry until I was starting at least 10 motors that size within the demand period, maybe not until I was starting 100.

If you're going to drag in steel mills again, realize that everything is in proportion. If I'm starting and stopping hundreds of motors, it's going on all the time and the very randomness makes any one particular motor insignificant. It's not like at 0700 hours we start ramping up the mill. It's a 24/7 operation.

 

[gadfly56]

Moderators, it looks like this thread is not updating properly.

 

[mivey]

Moderators, it looks like this thread is not updating properly.

It seems to have caught up. I also noticed T.M. moved to guest status. Is that related to the thread sticking? What is a guest status?

 

[iwire]

T.M.Haja Sahib has left the building.

 

[Besoeker]

T.M.Haja Sahib has left the building.

And elsewhere.
I hope nothing bad has happened to him.

 

[Electric-Light]

Most utility demand readings are based on a 15 or 30 minute sliding window and will not reflect the motor starting current.

This is correct. Billable peak is different from maximum draw.
Here is two definitions (because, each PoCo is different). Back in the days, it was basically giving into the capability of the meter, but with today's computerized meter, PoCo can setup its own rules in the program.

Monthly Demand: Monthly Demand is defined as the highest integrated kilowatt load during any 60-minute period occurring in the billing period for which the determination is made.

Definition of On-peak Demand: The maximum kW used by the customer in any fifteen-minute period between the hours of 10:00 a.m. and 10:00 p.m.
Let's look at the latter.

This means that during the 12 hour window shown, the highest average in any given 15 minute period will be used.
Imagine that you're driving your car. You're stuck in grid lock, then you see a clear path. You get frustrated, floor it, and get the car up to 110mph. Still, in 15 minutes you only traveled 5 miles, so the 110mph doesn't count as your peak. So, even though you got it up to 110mph, the 15min average "peak" is 20mph. (you can compare this peak with the motors coming on)
If you use an integration period of 3 seconds, you'll most likely capture the 110mph "max".

Later on in that day, you were driving on the freeway on cruise control @ 70mph and you sustained it for 30minutes. A 15 minute snap shot will show that "highest 15 minute average speed" as 70mph.

Suppose you were sizing the engine for that car. You figure that something that can get it up to 85mph based on the 15 minute average date. Nope, 110mph capability is needed. Even if this is only for one minute, if its required for one big critical piece of equipment, such as an elevator, it will fail to operate.

With electrical power, you don't necessarily want to document a very short surge, so you'd need use a peak meter with shorter integration period that is more in line with the holding duration of the flywheel effect of generator.

 

[Aramburo]

Relationship between Maximum demand and size of a diesel generator

It?s important to define how many hours operation.

There are differents level of diesel generator operation: Emergency, 8 hours; Prime, 12 hours; continuos, 24 hours.

The diesel generator capacity must be 20 percent over the maximun demand or peak demand , remember the LRA current about the starting of motors o bigger motor, diesel generator there is not the utility.