Instantaneous power consumtion

[russrmartin]

Hi guys. I'm trying to figure out how to calculate the instananeous power consumption here at the plant based on pulses counted off the meter. According to the power company, each pulse is worth 1.44 kW. Right now we're setting up a simple averaging system which we can track, but I'd like to take the data and calculate the instanaeous power consumption in any given 15 minute interval. The intervals are set by the power company. Anyone have any idea how I could do this? The idea behind this is to shut down machinesahead of time and realize the drop in usage rather than shutting them down and waiting to see what the drop is. Thanks.

Russ

 

[TConnolly]

According to the power company, each pulse is worth 1.44 kW.

Is that 1.44 kW, or 1.44 kWh? (I suspect the latter) because the two are very different things.)

Do you want to calculate kW or kWh?

 

[russrmartin]

Clarification

1.44 kW. We are billed in kWH. The pulses off the meter represent kW, which is then extrapolated out into 15 minute intervals. Those are then multiplied by 4 to get the kWH. What I need to do is reliably trend this data in these intervals to give us a display of what we are being billed for as consuming. We have an interuptible power agreement with the supplier now, and the last 2 days we have had to basically shutdown the plant due to the heat increasing power demands in the area. What I need is a reliable display of where we are running, so we can shutdown equipment accordingly and get below the set threshold by the power company. In a nutshell though, I just need to display an instantaneous rate of consumtion based on these pulses which come over time.

 

[bikerdude]

russ,

can't answer you rquestion but I thought I'd comment anyway. Our plant is on the power interruption program also. We have had to shut down 2 lines each day this week from noon until 8:00 pm. Here is my idea how to conserve energy during peak usage hours. Cut the air conditioning off to all the managers offices! Then thay could feel my pain. HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAH
BD

 

[kboggs]

Buy an IQ data plus II from Eaton Electric. or a metering module like that

 

[kamenges]

I'm with Alaric on this one. Power is by definition a rate. So to say that one pulse is equal to 1.44 kW doesn't seem to add up. Saying that one pulse per second is equal to 1.44 kW would make sense. In that case one pulse would equal 0.4 watt-hours of energy use. An easier way to think of this is that 3600 pulses equals 1.44 kW-H.

But I think more correctly that Alaric's original thought was correct. One pulse equals one kW-H. If this is the case you can convert back to rate by dividing by time. Specifically, divided 1.44 by the time between pulses in hours to get instantaneous power. If pulses are coming too quickly for that count a certain number of pulses and scale accordingly.

Keith

 

[Ken Moore]

Get a meter

I've been investigating meters for monitoring our power quality and usage, just about every meter I've looked at can do what you're asking for. I would call up your local switch gear supplier and discuss it with them, it might not be as expensive as a "home made" tracking system.

 

[Tom Jenkins]

The pulses from utility metering equipment are usually defined as kWh per pulse. They measure energy use. Totalling the number of pulses in a given period of time would give total energy use. To get kW you would divide the energy by the time interval between readings.

 

[RSL]

Using your existing setup you would have to do what Tom said. Count the pulses and divide by the sampling period to get kw. It will only be an average for the period. You cannot get actual instantaneous kw this way. It will always lag by the sample period and only be an average.

The time interval you mentioned is your demand period. Ours is also 15 minutes. The utility sets your peak usage surcharge based on the highest amount of kwh's used in any single period. They track our peak demand about 3 months out of the year. We pay the charge based on the peak set for the entire year though.

Ken is pointing you in the right direction if you really want good data to control peak load on the fly. We had a set of PT's and CT's installed by our utility company on our incoming power. We installed a power monitor attached to them whose data is avaliable via our Ethernet. We datalog, alarm and control based on realtime data from our meter.

We do get a demand synchronization pulse from our utility companies revenue meter so that we know exactly when our 15 minute interval begins and ends.

We have been able to identify several ways to control peak and save money. Our utility even approached us about what we did. We moved our peak to off peak hours. Since we use around 10 mw's it had an impact on their system too.

We use AB power monitors, but I agree all the big names will have a solution for your needs.

RSL

 

[russrmartin]

We are currently using the AB power monitoring system, and we are having problems, basically, the graph we have for our projected power demand for any current interval is not accurate at all. This is the reason for my desire to create my own code to reflect what our energy usage is. The pulses I am referring to are coming from our meter, and into the power monitoring system. If the pulses mean kW or KWH, matters none to me. I simply want an instantaneous graph of what our consumption is, no matter what the units are. I also do not want just an average calculation. This would be ineffective as a tool for us to use to get below our designated limit. What I really need to project is based on the average of our samples taken this period, what is the amount we will be billed for. The system AB setup for us really doesn't give us that. I cannot wait for a 15 minute period to be over before I get a final amount, as it will be too late at that point. I'll try get what is in my head out here so people can bang away at it.


1st pulse hits. = 1.44, 1 second into interval.
Average projected usage = 1.44 * 900 seconds(15 min.)= 1296

2nd pulse hits. = 2.88, 3 seconds into interval
Average projected usage = 2.88* 300 periods(15 min) = 864

And so on. Now, the only thing wrong with just displaying the 1296 and the 864 is that I only really need to pay attention to the 864. What I really want is to see the trend between the two, and exptrapolate that out over the rest of the 15 minute interval. This way, I can see that after 5 minutes into a period, we're shutting down power at a rate fast enough to get under our "limit" or not. For the data given, I'm at a slope of -216 per second. This would be a piece of cake to extrapolate. What I am confused about is how I handle say 285 of these points to create a accurate projection of where we are going to finish. Obviously, if we run steady at say 3000 for 12 minutes, then slope off at this -216, my projection should not say that I'll hit zero. Instead, it should be time weighted to show that even as I fall off at the end, I still ran high for the majority of the interval, hence my projection should say something like 2500 or something. Hopefully this is a little clearer now, and my apologies for being so long winded.

Russ

 

[RSL]

A few questions.

Which AB product are you using?

Is the device stand alone or are you using a PLC to calculate your projection?

How do you access the data?

What are you generating a graph with?

Are you getting a demand period synchronization pulse from your utilities meter?

Our meter does the calulation on the fly using realtime load data. You have to have CT's and PT's connected and be able to calculate real power that takes your power factor into consideration. A pulse per Kwh just wont get you as accurate as I think you want to be.

RSL

 

[russrmartin]

My apologies for the delay

I was fishing for the weekend, a break sure was welcome. Well, I have researched this further. #1, yes, we get a sync. pulse from the power company. #2. The value of our pulses is 1.44 KW for every change of state from our meter. It is indeed KW, not KWH. For every change of state from the meter, the total KW is summed. Our bill is essentially the sum of these changes of state(KW) divided by the interval time with a unit of seconds. This is then multiplied out to convert the seconds to hours, giving us the final KWH for the set interval which is 15 minutes. I created my own setup which basically does the same thing the power company is doing, but mine is much simpler. I gave up on trying to project what the energy for the interval would be. Instead, I simply look at the change of state of the meter. When this changes, I add the 1.44 to the previous total, then divide that total by the total interval elapsed time. With this calculation running all the time, the average interval KWH is updated about once a second, maybe a little more. This seems to give us a pretty good indication of where we are at thoughout the interval. It's not ideal, but I think it is workable. Thanks to all for the input.