Announcement

Collapse
No announcement yet.

Primary Transformer Circuit Breaker Tripping on in Rush

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Primary Transformer Circuit Breaker Tripping on in Rush

    Our telecom client has an inverter with an internal 40kVA, 480V delta to 120/208V wye transformer supplying a bypass circuit for the inverter. The only time the transformer has a load on it is when the inverter is in bypass. Normally DC power is converted to AC power to feed the critical loads. The client tests the generators every Thursday and about 25% of the time when the building returns from generator power to utility power, the primary circuit breaker trips.

    As I read it the NEC allows the primary protection to be sized for 250% of the transformer rated current if secondary protection is also provided (Table 450.3(B) which the inverter has. A 40 kVA transformer draws 48A on the primary side so the largest circuit breaker allowed by the NEC is 125A, the next standard circuit breaker size up from 120A (2.5*48).

    The panelboard where the 125A circuit breaker is installed is a Square D, NF Series, which will not accept a circuit breaker with adjustable time delay.

    Is there a commercial device available that could be added to the circuit to dampen the in rush current?

    #2
    180912-2057 EDT

    A thermistor. Found an 80 A at
    https://www.mouser.com/Circuit-Prote...iAAEgJwTfD_BwE
    You can search for higher current units, but possibly 80 A would work. Depends on load. Also could relay shunt the thermistor after a very short time, 100 milliseconds.

    A resistor shunted by a time delayed contact.

    .

    Comment


      #3
      Originally posted by Paul the BIG Finn View Post
      Our telecom client has an inverter with an internal 40kVA, 480V delta to 120/208V wye transformer supplying a bypass circuit for the inverter. The only time the transformer has a load on it is when the inverter is in bypass. Normally DC power is converted to AC power to feed the critical loads. The client tests the generators every Thursday and about 25% of the time when the building returns from generator power to utility power, the primary circuit breaker trips.

      As I read it the NEC allows the primary protection to be sized for 250% of the transformer rated current if secondary protection is also provided (Table 450.3(B) which the inverter has. A 40 kVA transformer draws 48A on the primary side so the largest circuit breaker allowed by the NEC is 125A, the next standard circuit breaker size up from 120A (2.5*48).

      The panelboard where the 125A circuit breaker is installed is a Square D, NF Series, which will not accept a circuit breaker with adjustable time delay.

      Is there a commercial device available that could be added to the circuit to dampen the in rush current?
      I would guess that this is more related the sources being out of phase when the transfer occurs. Do you have the option of an in phase monitor for the ATS? Or if you tolerate it just program a delay in transfer from genset to utility.

      Comment


        #4
        It has to do with the inrush, which is not going to be fixed by a higher rated breaker, per se. Since it only trips 25% of the time most likely what is happening is that the primary winding is being connected at around the zero crossing of the primary voltage. This can cause extremely high inrush currents that can last for a few seconds. A device with an adjustable trip setting is going to be one solution.
        "Just because you're paranoid, doesn't mean they're not out to get you"

        Comment


          #5
          Originally posted by kingpb View Post
          It has to do with the inrush, which is not going to be fixed by a higher rated breaker, per se. Since it only trips 25% of the time most likely what is happening is that the primary winding is being connected at around the zero crossing of the primary voltage. This can cause extremely high inrush currents that can last for a few seconds. A device with an adjustable trip setting is going to be one solution.
          I thought that the residual flux and zero-crossing-enhanced inrush current lasted for only a few cycles of 50/60 Hertz, not for a few seconds. Do you have a reference I should look at?

          Comment


            #6
            Originally posted by GoldDigger View Post
            I thought that the residual flux and zero-crossing-enhanced inrush current lasted for only a few cycles of 50/60 Hertz, not for a few seconds. Do you have a reference I should look at?
            That was supposed to say a few "cycles"; good catch
            "Just because you're paranoid, doesn't mean they're not out to get you"

            Comment


              #7
              I presume that the transfer from generator to utility is 'open' since the transformer is being re-energized.

              What is the transition time? Could the delay be increased?

              My thought is that the inrush is being made particularly bad because there is still lot of residual flux in the core, getting added to the normal problems caused by timing of the AC cycle when the transformer is energized.

              -Jon

              Comment


                #8
                Inrush

                Transformer inrush is quite sensitive to the timing of energization on the cycle. But is also affected by the residual magnetism in the core when being energized. If you can get a thermal-only circuit breaker, no magnetic element, you still provide overcurrent protection as required but will avoid tripping on inrush.
                e^(i pi) = -1

                Comment


                  #9
                  Originally posted by beanland View Post
                  Transformer inrush is quite sensitive to the timing of energization on the cycle. But is also affected by the residual magnetism in the core when being energized. If you can get a thermal-only circuit breaker, no magnetic element, you still provide overcurrent protection as required but will avoid tripping on inrush.
                  Or, you could tweak the short time setting to the highest possible setting --> if your breaker has an electronic trip unit.
                  @Paul,
                  What type of breaker do you have?

                  Comment


                    #10
                    Originally posted by topgone View Post
                    Or, you could tweak the short time setting to the highest possible setting --> if your breaker has an electronic trip unit.
                    @Paul,
                    What type of breaker do you have?
                    I believe he said Square D NF breaker -not a lot of options for them, basically like the QO line but rated 277/480 instead of 120/240. Also fairly sure that 125 amp is the highest amp rating available for this particular line. If it won't hold one needs to either find a way to lessen the inrush or use some other line of breaker
                    I live for today, I'm just a day behind.

                    Comment


                      #11
                      Paul (tBF)...

                      1) Since the Xfmr is already energized before load is applied, then, discussion about residual-magnetism should can be ignored!

                      2) The simplest solution, if the nature of the load is known, is to set the protective device's tripping-point to about 2x the load's inrush. The inrush-current can be easily determined wiith a clamp-on!

                      3) A more complex solution is to sync the inverter to the Xfmr!

                      Regard, Phil Corso!

                      Comment


                        #12
                        Originally posted by Phil Corso View Post
                        Paul (tBF)...

                        1) Since the Xfmr is already energized before load is applied, then, discussion about residual-magnetism should can be ignored!

                        2) The simplest solution, if the nature of the load is known, is to set the protective device's tripping-point to about 2x the load's inrush. The inrush-current can be easily determined wiith a clamp-on!

                        3) A more complex solution is to sync the inverter to the Xfmr!

                        Regard, Phil Corso!
                        Phil, the way I understand the description the ATS that switches between POCO and generator is connected to the primary of the bypass transformer. That means that residual flux in the primary is indeed a factor unless the generator syncs to POCO before the reverse transfer.
                        The inverter is supplying power to the load continuously through both transfers as well as any momentary outage and so its synchronization or lack of it is not an issue.
                        I suspect that the problem occurs only during the reverse transfer because there is a delay before the generator picks up the load at the beginning of the test and a large portion of the instantaneous flux has decayed. A simple time delay with the generator turned off before the back transfer to POCO should, I believe, resolve the problem.

                        Sent from my XT1585 using Tapatalk

                        Comment


                          #13
                          Breaker Frame Type

                          Paul
                          You said the tripping breaker is in a NF series panelboard.
                          What is the breaker frame? EDB, EGB, EJB etc........?
                          Can you remove the panel cover and take a closeup pic of breaker front?
                          We need to find out if the inst (mag trip) setting is fixed or adjustable and
                          what that value is.
                          Ifyoucan'texplain itsimply youdon'tunderrstanditwellenough- Albert Einstein

                          Comment


                            #14
                            GoldDigger...

                            Your interpretation is correct. I misinterpreted the layout!

                            However, in my 64 year career, I always recommended the utility-fed Xfmr be continuously energized, so that load transfer time-delay was limited to just that of the ATS!

                            Phil

                            Comment


                              #15
                              Better still, a sketch or a diagram could be of great help in the analyses here.

                              Comment

                              Working...
                              X