Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

KVA jump with KW decrease?

  1. Oct 2, 2013 #1
    Hi Guys,

    So we have a melting furnace that is melting Aluminium. After 2 days of holding a melt to see how the furnace performs in a holding situation, I changed the PID control to have a longer sampling time and hence smoother control. What I noticed was the KW seemed to drop, although the average usage was probably the same. The KVA value had a definite increas in the average value. I can post a graph of the results if anyone is interested?

    Can this be explained? The elements a SiC and are controlled by a thyristor card by current setting.

    Thanks for any help.
     
  2. jcsd
  3. Oct 2, 2013 #2
    Do post your results.
     
  4. Oct 2, 2013 #3
    Graph

    This is using an Elster 3 phase meter, integrating over 5 minutes. The KVAh term is just me looking at what the KVA has been doing over the previous hour.
     

    Attached Files:

  5. Oct 2, 2013 #4
    You can see the kW has stabalised, but using the same average. The KVA value however at the exact time I changed the PID sampling time shot up.

    I can only think that because the drive is now staying on constantly and never turning off there is constant switching to keep the power at 20%. Before it could turn on to 100% where it uses the power more efficiently and then off again, meaning the capacitance introcuded into the system was not as great? I'm not really sure though.
     
  6. Oct 2, 2013 #5

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    What is "PID sampling time?"

    Your graph looks like a classic case of closed loop oscillation from way too much gain or time delay...

    If oscillation is around 2/3 of integral time try lengthening that setting
    if it's faster try less proportional gain
    try switching derivative all the way out and add back as required to reduce overshoot in response to step change of setpoint

    Do some reading on control system tuning. Old line companies like Fisher Controls and Bailey Meter pretty much wrote the books.
     
  7. Oct 3, 2013 #6
    Hi Jim. Thanks for the reponse. Unfortunately, that was not the answer I was looking for. As it's an aluminium melting furnace, with 2 tons of Aluminium inside, I believe the time delay factor is out of my hands, as it honestly takes ages for even full power to adjust the temperature ever so slightly.
    My question was actually on the reasoning behind the KVA jump when the system had changed to a larger sampling time and more stable?
     
  8. Oct 8, 2013 #7

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    Are the thyrisistors fired at zero crossing, ie on and off for interger numbers of cycles, or are they phase angle controlled?

    I suspect it's a matter of metering - how does that Elster meter work?

    reason i ask is

    from http://nrldc.in/Commercial/calculations/SEM_Hand_book.pdf [Broken]
    which i got from a search on Elster

    is this it?
    http://www.elstersolutions.com/assets/products/products_elster_files/A1700_31052012.pdf
     
    Last edited by a moderator: May 6, 2017
  9. Oct 8, 2013 #8
    Hi Jim,

    Yes it's Phase Angle Controlled and I am using the A1140 meter.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook