Determining existing load for load addition

[philly]

I'm looking at a situation where I need to add a 780A load to an existing 480V Switchboard to feed new chiller loads (780A is calculated new chiller loads). The existing 480V Switchboard is a 1600A rated Switchboards with (3) 800A feeder breakers feeding downstream panels.

I'm trying to determine what the existing load is on this 1600A Switchboard to see if there is enough capacity to add the new chiller loads. The (3) 800A feeder breakrs feed downstream 800A ATS's which in turn feed 800A emergency panels.

I've had not luck tracking down any sort of load schedules for either the 1600A Switchboard or downstream 800A panelboards to give an indication of the existing calculated load. I have the one-lines that show all the feeders from these downstream panels (feeding other panels and loads) but nothing that quantifies loading.

The other option in lieu of any load data would be setting up power monitors at a location to be able to measure the existing loads. Unfortunately the 1600A Switchboard in question is an emergency board and would only have load during emergency gen conditions so we'd likely have to setup (3) meters at the load side of each ATS or the panelboards themselves and then have to manually add loads to quantify total loads.

I'm trying to come up with ideas on how to potentially quantify the existing loads without having to go through the process of metering. I'm curious to hear any ideas how others would approach this?

 

[philly]

I'm curious if there is an accepted method from estimating existing load based on breaker sizes in a panel. Obviously you can make the assumption that feeder/branch breakers are sized based on continuous load (125%) so taking 80% of these ratings may get you in ballpark of calculated load. This approach always seems to be very conservative and usually still leads to a values greater than the rating of the panel itself.

I'm curious if anyone has ever used this method or other similar methods in lieu of calculated load data or metering?

 

[JoeStillman]

There is no "accepted method" to determine loads given nothing but breaker sizes. Connected load is related to breaker size, but demand load is not. Demand loads are only calculable from measurement data. I always ask for a year of electric bills. If this is not the only gear on the electric bill that won't work for you.

Ammeter readings are about the only thing you can hang your hat on.

 

[philly]

There is no "accepted method" to determine loads given nothing but breaker sizes. Connected load is related to breaker size, but demand load is not. Demand loads are only calculable from measurement data. I always ask for a year of electric bills. If this is not the only gear on the electric bill that won't work for you.

Ammeter readings are about the only thing you can hang your hat on.

There is a power meter (ASCO 5200 series) on each of the (3) ATS's that provide current readings however these only provide real-time instantaneous readings. I don't know how practical these can be for evaluating existing load seeing that they are instantaneous and readings depend on load provide etc... at that exact moment.

The meters do however have a max KW demand feature that stores the max kW Demand value in memory since the meter was last powered-up or reset. I don't know again that I'd consider this an effective means of evaluating existing load given the fact that there is no PF associated so we'd have to guestimate PF and Voltage to determine current demand. Curious to hear others thoughts on using this max KW demand value as means for evaluating existing load provide or if not at least to provide initial value for consideration prior to 30-day metering process?

 

[drktmplr12]

Existing load demand data needs to comply with 220.87. If 1-year data is not available, the exception is "maximum demand (measured of average power demand over 15 minute period) for a 30-day period using a recording ammeter or power meter connected to the highest loaded phase of the feeder or service."

The max kW demand feature can be used as a reality check, but cannot be used to calculate the load based on the NEC. Often times we face this issue where we are certain the diversity of the system means the transformer will never be overloaded but we can't prove it on paper so we have to place our meters and wait 30 days. You will often see engineers apply 80% demand factor on everything. While this is practical, there is nothing in the NEC that permits doing so.

 

[philly]

Existing load demand data needs to comply with 220.87. If 1-year data is not available, the exception is "maximum demand (measured of average power demand over 15 minute period) for a 30-day period using a recording ammeter or power meter connected to the highest loaded phase of the feeder or service."

I take the part in bold to mean that when the meter takes 15 min window snapshots over the course of the metering period we are only concerned with the "maximum average" value from all the samples taken (number of samples depends on sampling rate) and not the just the "maximum value". The maximum value during that window will be the absolute maximum instantaneous values recorded that window which could represent short duration transients from motor starting etc...? The maximum average value is an average of all the maximum values and better represents maximum steady stage loading.

The max kW demand feature can be used as a reality check, but cannot be used to calculate the load based on the NEC. Often times we face this issue where we are certain the diversity of the system means the transformer will never be overloaded but we can't prove it on paper so we have to place our meters and wait 30 days. You will often see engineers apply 80% demand factor on everything. While this is practical, there is nothing in the NEC that permits doing so.

Yes this is the situation we're in on this one where we believe there is diversity but need to meter to prove and document. I agree the kW can provide a reality check. What would you assume as a pF to use with this initial max demand gut check not knowing the nature of the loads?

 

[drktmplr12]

I take the part in bold to mean that when the meter takes 15 min window snapshots over the course of the metering period we are only concerned with the "maximum average" value from all the samples taken (number of samples depends on sampling rate) and not the just the "maximum value". The maximum value during that window will be the absolute maximum instantaneous values recorded that window which could represent short duration transients from motor starting etc...? The maximum average value is an average of all the maximum values and better represents maximum steady stage loading.

That is my understanding as well.

Yes this is the situation we're in on this one where we believe there is diversity but need to meter to prove and document. I agree the kW can provide a reality check. What would you assume as a pF to use with this initial max demand gut check not knowing the nature of the loads?

0.8

one final option would be to do the calc as though it was new. depends on your experience with the type of facility/building.