| Arc Flash Forum https://brainfiller.com/arcflashforum/ |
|
| Label sub-panel with main panel hazard? https://brainfiller.com/arcflashforum/viewtopic.php?f=24&t=1085 |
Page 1 of 1 |
| Author: | arcflash71 [ Tue Apr 27, 2010 2:56 pm ] |
| Post subject: | Label sub-panel with main panel hazard? |
Hello, It has been difficult to obtain complete data for some of my facilities where there are ongoing system studies for arc flash. In one particular case I have no make/model for load breakers inside of 480V motor control centers. We require that load disconnects and sub panels be labeled, yet I cannot obtain an outage to obtain the information. Our design documentation is poor and the information is not available through that route. Labeling load disconnects and sub panels with the motor control center HRC appears to be a conservative approach. My reasoning is that in order for the hazard to increase as you go "down" in the system, the fault level would have to fall out of the INST trip region of the load breaker. Due to voltage drop considerations that seems unlikely if the breaker was sized appropriately during design. What are your thoughts? How would/do you handle this situation? Assuming my reasoning is correct, would you consider my proposed approach risky? Thanks |
|
| Author: | acobb [ Tue Apr 27, 2010 5:16 pm ] |
| Post subject: | |
I would say that voltage drop is not the issue.....it is the additional impedance of the circuit conductors. My approach is to label a device as worst case. If a hazard does not exist on the line side of the equipment, then label it for the load side IE. If a hazard exists on the line side of the equipment, even while intending to work on the load side, then I believe that you need to label it worst case for either the device being worked or the upstream device clearing time, whichever is worse. Who knows where the arc might be initiated. |
|
| Author: | arcflash71 [ Wed Apr 28, 2010 8:47 am ] |
| Post subject: | |
I didn't make my question clear. You're right that voltage drop is not the issue; it's my explanation for why small cables (high impedance) would NOT be spec'ed for long runs of cable. The reasoning was that cables would be spec'ed to prevent voltage drop, thus allowing a fault level adequate for fast, selective clearing of the fault. Additionally, this is not a question about line/load side as with the main breaker in an MCC. This is a question about how the hazard level changes as you move "down" in the system through coordinated series overcurrent protective devices. For the purposes of my question each faulted bus is separated from the other faulted buses by enough cable to eliminate the possibility of fault propagation to the line side primary OCD. We are in agreement that the "worst case" should be labeled. The question is, in a properly designed radial low voltage system with inverse-time breaker trip curves, does the hazard always decrease as you move from one panel or MCC to a subpanel or load disconnect? Thanks for your comments |
|
| Author: | JBD [ Wed Apr 28, 2010 10:28 am ] |
| Post subject: | |
arcflash71 wrote: The question is, in a properly designed radial low voltage system with inverse-time breaker trip curves, does the hazard always decrease as you move from one panel or MCC to a subpanel or load disconnect?
Always? - NO. My company uses an energy boundary method for determining a maximum (and sometimes minimum) amount of conductor needed between an MCC bucket and the downstream disconnect. We use the available SCA at the MCC (or for simplicity the lowest SCA in the facility) and combinations of starter protective devices and conductor sizes. The result is a table of distance versus HRC, for example: Given 7.55 SCA @ 480V, Protective device HMCP-150A with #2AWG: Never less than HRC1, HRC=1 upto 465ft, HRC=2 upto 519ft, and never exceeds Cat. 3 However the results are only valid for these specific combinations, so before a label can be installed the MCC bucket must be opened and the components confirmed. |
|
| Author: | JJH [ Wed Apr 28, 2010 10:35 am ] |
| Post subject: | |
I have seen cases where the IE on downstream equipment went up due to the impedance of the cable even when the cable was properly sized. Sometimes the fault current at your main panel is low to start with, so as you go down your system you fall out of the instantenous ranges of your breakers. I have also found some breakers without labels. I ignore these breakers and use the next upstream device as the clearing device. It makes for an artifically high IE at the downstream device, but it is better than someone getting hurt because I chose the wrong breaker to model. In my recommendations, I then recommend to replace these breakers to bring down the IE. |
|
| Author: | arcflash71 [ Wed Apr 28, 2010 12:46 pm ] |
| Post subject: | |
JJH and JBD, Thank you both for your informative comments. I would be curious to know under what conditions you have seen IE on downstream equipment increase due to impedance of the cable. These are the situations I can come up with based on your comments and my own guesses: 1) Feeder breaker improperly sized to protect cable. 2) Feeder breaker without an INST trip element. 3) Very low available fault. 4) Very long cable run. Am I missing anything obvious? Regards |
|
| Author: | JJH [ Thu Apr 29, 2010 10:42 am ] |
| Post subject: | |
I have run across all of those situations. I guess my point is that until you gather the data you don't know if the cable is properly sized, or if the inst. trip is set at 5x or 12x, or what the fault current is at the downstream device, or low long the cable run is, or you could have a combination of factors (inst. trip set at 12x with a moderate fault current available with a medium length cable run). I don't label anything until the data has been gathered and analyzed. Too many variables. |
|
| Author: | arcflash71 [ Thu Apr 29, 2010 11:10 am ] |
| Post subject: | |
JJH, Your point is well taken. My goal is to identify cases where conservative assumptions can be made to help the process stay on track. As a company our facilities are huge and widespread, our wants are virtually unlimited (perfection) while our resources are highly constrained. It's like surfing a landslide in that the details stack up and bury you with the slightest misstep. To the extent that "perfection is the enemy of good enough" I'm balancing risk of delay against the risk of injury due to some corner case hazard. In other words I'm focusing on what matters. Of course the hard part is knowing what matters, proving your approach is sound and then documenting your approach... Based in part on this conversation I believe that the right approach for my organization is to determine the hazard based on the clearing time of the next (known) upstream OCD. For cases where not even the cable size/run is known I guess I'll wait for data... Thanks again for the feedback. Regards |
|
| Page 1 of 1 | All times are UTC - 7 hours |
| Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |
|