IET
Decrease font size
Increase font size
Topic Title: ARC Flash Incident on Ship
Topic Summary: Cause?
Created On: 25 December 2011 08:23 PM
Status: Post and Reply
Linear : Threading : Single : Branch
1 2 Next Last unread
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 25 December 2011 08:23 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Merry Christmas guys - hoping you can perhaps share any experience you may have on arc flash incidents.

We have recently experienced the catastrophic failure of 2 - ABB TMax T5 500A MCCBs when operating the switching mechanism.

http://www05.abb.com/global/sc...le/1sdc210003b0201.pdf

The circuit breaker is on the 690V switchboard which feeds a lenght of cabling then the incommer breaker of an ABB drive unit. This then feed a single motor.

The first incident was caused following a trip of the circuit breaker during operations. The electrician investigated, found the circuit breaker in the trip position and no fault so decided to reset the trip on the breaker by pushing the lever down from the tripped position to the open position. He reported a strange noise coming from the breaker so stood to the side when switching.

This action of resetting the trip (pushing lever from trip to open position) caused the circuit breaker to explode catastrophically causing a complete flash over of the unit and causing significant injury to the electrician. It was reported that the downstream breaker of the circuit being energised (drive incommer breaker) was open when this happened.

The switchboard housing this breaker is 690V. This is fed from the 6.6kV network and is onboard a ship. The system is a high impedance earthed system as is common on most "High Voltage" ships. The neutral point of each generator is connected to the ships hull by a resistor. As is the case with most ships - no neutral exists.

The second incident came about after the burnt out breaker was replaced and the switchboard cleaned up. The assumption following the first incident was that the first breaker had developed a fault.
A new breaker was delivered from ABB and the vessel electrician installed it. Worried about the possible re-occurence of the first incident the electrician made sure and checked the breaker downstream at the drive was open and the fuses for the pre charge circuit were open. Effectively the breaker was being closed onto the cables only. The electrician meggered the cables between the breaker in the switchboard and the drive and found no fault to exist. Prior to energising the breaker both electricians decided to wear the arc flash protection gear kept on board for the 6.6kV system. Wearing some of the arc flash suit (hood and gloves) the breaker was closed when it subsequently failed a second time. This time the failure was a lot more severe and both men, although requiring significant treatment for burns, are very lucky to have been wearing the suitable PPE not to be more seriously injured.

We have spent all week examining the accident site and can find no reason for this incident. The drive has no fault - confirmed by ABB drive Engineer, neither does the cable or the switchboard. It is also worth nothing that the point where the breaker failed was not at the contacts but at the internal protection device below the operating lever on both occasions. This houses 3 CTs and a small electronic tripping device. These breakers are very simple to install out the box therefore an installation fault may be possible unlikely. The first breaker was purchased in 2008 during build and the second breaker was bought from ABB last week in Houston.

A cursory glance over the engineering / protection study / short circuit calc shows all this equipment to be suitably rated for this duty.

No back feeding was possible as the breaker which failed, was only feeding a single circuit downstream.

We have meggered the cables using a 2kv megger to establish if there is a short on the downstream cable but the results come back clear - as they did prior to the first incident. We have also checked continuity of downstream cabling. All good.

No point of damage or flashover appears in the switchboard and everything looks fault free downstream.

We really need to find the root cause of this problem to make sure this does not happen again but everyone involved in struggling with ideas - even sending this further afield to colleagues and other Engineers has all met with the same response.

Hopefully you guys have some experiences and suggestions you can share to ensure we can put the system back to work with the confidence no one else will be injured.

Thanks and Merry Christmas

Mike
 26 December 2011 09:03 AM
User is offline View Users Profile Print this message



sparkyali

Posts: 7
Joined: 26 October 2007

Hi mike,
Reading your email was very strange - sounds impossible! No wonder it's causing problems,
You said the blow up both times was the trip unit under the breaker housing ct's and trip coils.
I'm not sure if you have external connections to this part of the breaker for remote monitoring or remote tripping, if so this could be the cause. A different fault within the switchboard could be sending the wrong voltage to the trip coils, or even open circuit ct's can go bang if they are used remotely. Hope you post the results when you find them - good luck!!
If it's just a stand alone breaker I can't suggest something you have not already done.
Good luck and merry Xmas hope you post the outcome when you find it.
 26 December 2011 12:35 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Hi,

Merry Christmas

Thanks for the reponse.

The CTs are completely internal to the breker - the breaker has a small electronic trip device at the bottom with come dip switches and the CTs just provide an input to that OC and SC tripping.

The only external connections - one is for a UV coil supply to trip the breaker. The second has also some remote indication - open/closed/tripped for the automation system. No external CT input / output.

We have given the cables a full on megger test and can find no fault.

Do you know if any more complex ways exist to test cables. For a gennerator / motor winding I know of the Baker test unit - would something a bit more hi tech such as this be available for LV cables.

The only other piece of informarion which may be relevant is that on the back of the cradle of the downstream circuit breaker on the incomming side there are 3 capacitors which tie the live side to ground. These are very small and appear to be there as some kind of grounding. The cables and capacitors are very small (1.5mm cable) and upon examination one did show slight signs of burning although we suspect this was caused during the megger testing. We are certain these capacitors which link each live phase to earth have not caused a current of the size we require to do this damage - they are all still in tact.

Thanks

Mike
 26 December 2011 05:08 PM
User is offline View Users Profile Print this message



ArthurHall

Posts: 735
Joined: 25 July 2008

Is the 690V system and earth free IT system? If so is there an earth fault up stream?
If this is the case the full 690V could be between one phase and earth. Is the protection unit rated for this or just 690V phase to phase.
 26 December 2011 05:54 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Arthur - thanks for the input.

We have got some degree of earth fault repoted throughout the vessel but nothing that would be a full on problem.

The vessel has got earth fault protection on the 6kV side through means of a neutran earthing resistor and symap generaor protection relay. It is set so a full on earth on the 6.6kV side would trip the breaker.

Say we did have a full earth on one phase to earth - this would mean that the phase voltage on one phase to earth would be zero and on the other 2phases the phase voltages would be 690V (same as the line voltages). Is this what you are suggesting?

Not a bad theory, I will revert with the results of the testing of the upstream system. They were just disconnecting the secondary sode of the transformer (a LOT of cables) yesterday.

 26 December 2011 07:16 PM
User is offline View Users Profile Print this message



ArthurHall

Posts: 735
Joined: 25 July 2008

You say you have a 690V system yet ABB say the breaker is rated for 600V, have you checked the breaker is suitable for 690V?
For a 690V system you have about 398V phase to earth but if one phase is earthed you will have 690V phase to earth which could be too much for the protection unit.
If the breaker is rated for 600V it will be expecting about 346V to earth approx half of 690V.
 26 December 2011 11:57 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Thanksd Arthur - the breaker is a 690V 3 pole breaker which I assume is 690V phase / phase.

Interesting question - I will let you know if we uncover anything testing the upstream circuit.

Best wishes

Mike
 27 December 2011 12:06 PM
User is offline View Users Profile Print this message



sparkyali

Posts: 7
Joined: 26 October 2007

hi mike,
sounds like the ct's/ trip coils are not to blame, but regarding the capacitors on each phase, you mentioned the circuit was feeding a ABB drive. i've sometime seen people install such capacitors on the upstream side of drives on ships in a desperate attempt to clean up the waveform destroyed by the drive. this also leads to another possibility - the drives aften are fitted with inductive chokes on the upstream side again to protect against waveform damage. it could be there is a inductive choke in the circuit out of sight somewhere that is blown/shorted. if it has a soft start on it you may not have seen it even with a megger test.
regards
ali
 27 December 2011 12:37 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Thanks Ali.

im heading back out to the ship on Thu so will have another look then. on the drawings the only thing shown are the 3 small capacitors which as you say are a crude attempt to clean up the waveform.

I will have a look and revert to see if there is a choke - there is nothing in the drawings but as experience tells me that means nothing..

Mike
 27 December 2011 05:13 PM
User is offline View Users Profile Print this message


Avatar for dougflorence                                      .
dougflorence

Posts: 74
Joined: 25 July 2008

Maybe I am being naive but unless the breaker is being exposed to some very unusual fault when it is switched on, surely it would not be expected to explode, even if switched into a dead short. I would have thought that almost by definition it ought to break the circuit again without hazard to the person switching it. Have ABB expressed any interest in finding out what the problem is? I would have thought that as responsible manufacturers, they should be extremely concerned if one of their breakers has caused injuries when it appears to be being used in a correct fashion.

You say the explosion occurred in the auxiliary box below the breaker handle. Surely the circuits connected to that are where you need to focus your attention. Is there enough left of the remains of the blown breakers to determine in which components the energy release occurred?
 28 December 2011 12:28 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Thanks Doug - yes we have sent all the bits to ABB and have asked for their conclusions - so far nothing and I am not sure they will provide anything conclusive.

I agree regarding the magnatude of the fault and that the breaker should still be in tact.

THe points at which the breaker connects to the busbars are totally clean - no sign of a fault. It seems to have shorted out internally in the protection circuit.

Thanks

Mike
 30 December 2011 03:25 PM
User is offline View Users Profile Print this message



williamjohn

Posts: 178
Joined: 22 November 2010

Hello Michael
This is probabely a stupid suggestion but has the 6.6kv supply been inadvertantly connected to the 690 volt breaker, perhaps by a short in the transformer.

If the 690 volt breaker is of the correct MVA fault rating for the system, it should not explode whatever fault occurs unless it has been inadvertantly connected to a higher fault power system or a higher voltage system.

Best wishes and good luck
John

Edited: 30 December 2011 at 04:38 PM by williamjohn
 30 December 2011 04:37 PM
User is offline View Users Profile Print this message



ArthurHall

Posts: 735
Joined: 25 July 2008

Mike
Have you established yet if the protection unit was rated at 690V to earth?
 30 December 2011 11:59 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Arthur- we are still waiting on ABB's formal response.

The more I consider this the more to me it does not seem to be involved with current. A lot of things point to the cause of this being a flashover caused by excessive voltages - namely:

1. No short or fault can be found out with the damaged zone in the breaker.

2. The breaker was open on both occasions when the fault happened - no current flow through the breaker. Verified both times by Megger testing the downstream side of the breaker).

3. Other issues and small tell tales signs of arcing at the back of the downstream breaker (drive incomer) This shows some signs of shorting out or arcing but not enough to sustain the required fault current. We notice some scorching between phase and a small amount of metallic "spatter". Nothing excessive and discounted by the drive service Engineer who attended as insignificant.

4. We have also had reports that the transformer was recently tapped down due to excessive voltages burning out control cards in some circuits.


Any thoughts?

The more I think about it nothing seems to point to current.

Also what could cause this problem apart from a primary/secondary short? Harmonics? Capacitive switching?

Thanks for the help so far

Mike
 31 December 2011 09:20 AM
User is offline View Users Profile Print this message



williamjohn

Posts: 178
Joined: 22 November 2010

Michael
I can only make sense of what has happened if there is a fault between the 6.6kv winding and the 690 volt winding AND for the 690 volt system neutral to be earthed downstream of the 690 volt breaker. Then when the transformer is energised, the 690 volt breaker becomes energised at 3.8 kv to earth. This could cause corona in the breaker, being the strange noise heard.

Now when the 690 volt breaker is closed, it puts an earth fault on the 6.6 kv system which passes through the 690 volt breaker and it tries to open. With the 690 volt breaker trying to break a 3.8 kv fault at much higher MVA than could normally occur on the 690 volt system, the breaker could well explode.

Look for a fault between the primary and secondary windings of the transformer.

Good luck
John
Edit 31 Dec. You say that the 690 volt system neutral is not earthed. However it may be that clearances are less in the breaker when closed so a 3.8 kv flashover may perhaps occur within the breaker when it is closed. All the evidence points to a fault between primary and secondary windings in the transformer.

Edited: 31 December 2011 at 05:32 PM by williamjohn
 31 December 2011 09:24 AM
User is offline View Users Profile Print this message


Avatar for sfchew.
sfchew

Posts: 589
Joined: 10 December 2002

Dear Mike,

With more clarification we will eventually find the cause.

"The first incident was caused following a trip of the circuit breaker during operations. The electrician investigated, found the circuit breaker in the trip position and no fault so decided to reset the trip on the breaker by pushing the lever down from the tripped position to the open position. He reported a strange noise coming from the breaker so stood to the side when switching.

This action of resetting the trip (pushing lever from trip to open position) caused the circuit breaker to explode catastrophically causing a complete flash over of the unit and causing significant injury to the electrician. "

Some points to clarify.
1. Was the noise heard when breaker was being reset?
2. Was the explosion caused when the breaker was being reset?
3. Was there any attempt to switch on the breaker?
4. Was there any upstream tripping during the explosion?

"The second incident came about after the burnt out breaker was replaced and the switchboard cleaned up. The assumption following the first incident was that the first breaker had developed a fault.
A new breaker was delivered from ABB and the vessel electrician installed it. Worried about the possible re-occurrence of the first incident the electrician made sure and checked the breaker downstream at the drive was open and the fuses for the pre charge circuit were open. Effectively the breaker was being closed onto the cables only. The electrician meggered the cables between the breaker in the switchboard and the drive and found no fault to exist. Prior to energising the breaker both electricians decided to wear the arc flash protection gear kept on board for the 6.6kV system. Wearing some of the arc flash suit (hood and gloves) the breaker was closed when it subsequently failed a second time. This time the failure was a lot more severe and both men, although requiring significant treatment for burns, are very lucky to have been wearing the suitable PPE not to be more seriously injured. "

Some points to clarify.
1. Was there any upstream tripping during the explosion?
2. Was there any fault indication on the breaker?
3. What is the total generators capacity?

"We have also had reports that the transformer was recently tapped down due to excessive voltages burning out control cards in some circuits."

What was the value of excessive voltage and voltage noticed during the incidents?

Please make a measurement on the phase - earth voltage.

"Other issues and small tell tales signs of arcing at the back of the downstream breaker (drive incomer) This shows some signs of shorting out or arcing but not enough to sustain the required fault current. We notice some scorching between phase and a small amount of metallic "spatter". Nothing excessive and discounted by the drive service Engineer who attended as insignificant. "

The above is worth investigating.

Regards
Chris Chew

What is the length of the cables?
 31 December 2011 10:00 AM
User is offline View Users Profile Print this message



dlane

Posts: 681
Joined: 28 September 2007

Hi Mike,

I am sorry to hear of your incident, my regards to your electricians and their families and I wish them a full and speedy recover from their injuries, especially at this time of year.

Most of my knowledge of arc flash incidents is on land based systems and I am not too familiar with ship installations so I apologise if I mention something not appropriate to your systems.

Arcing faults are different to normal system faults in that the actual fault is of a higher impedance than a short circuit fault, consequently the arcing current of the fault is much lower than the short circuit fault current and the protective devices take much longer to operate. For a 690V system I would have expected initial direct contact but that would then be blown away by the arc as it is initiated and then sustains itself so the initial short may not be detected post event.

You say that your 6.6kV system is impedance earthed, is this also the case for the 690V system? In my experience of these systems the impedance limits the current to a value below the arcing current so an arcing current cannot be sustained unless the impedance / resistor is faulty. If the 690V system is impedance earthed then only a phase to phase fault can generate enough energy to sustain an arc so this is where I would concentrate my investigation.

John's question of a transformer fault is an interesting possibility. If the secondary voltage is excessive, the breaker may be able to withstand it initially but operation of it may change clearances within it and initiate the arc. Your initial incident where a strange noise was heard may be indicative of this happening. Arcing faults act in strange ways that we cannot predict and will initiate in one part of a breaker but actually exit at another point due to a weakness found in the construction of them. Along these lines, has the tapchange been carried out correctly? A link based system may be more problematic than a tapchange switch, or is there a fault in one tap postion that is not present in others? . A turns ratio and insulation tests on the transformer may be prudent.

As an aside point you have local generation, this may also exacerbate your fault conditions, generators go through transient and subtransient fault conditions where the energy levels are much higher before they reach the steady state conditions, a breaker operating in the transient and subtransient stages would need a much higher breaking capacity. This obviously depends on your system configuration and the type of prime mover for your generator which I have no knowledge of.

Just a few words on safety. If the PPE has been specified correctly to match the arc energy levels then the wearer has a 50% chance of sustaining a second degree burn. Quite rightly you don't mention the extent of the injuries but I get the feeling they may have been greater than this? The medics should be able to advise if the injuries are from the initial blast or from secondary burning of clothes. Burn injuries can still be suffered if incorrect clothing is worn underneath flame / arc resistant PPE.

Although arc resistant PPE does have its place I am not a big fan of it and much prefer engineering solutions and below are 3 website links for arc protection systems and remote breaker operation devices.

Vamp arc protection

ABB arc protection

I have utilised both of these systems on installations, my personal preference is for Vamp as I find it more configurable than the other systems on the market but they all work in the same manner. They can be awkward to install but do protect personnel and limit switchgear damage. There are other systems on the market (UTU, Camto, Klockner Moeller) but I have not installed these.

Using these systems allows lighter PPE to be worn and operators can walk away injury free from an arc flash incident.

Breaker remote operation

The remote operators may be of use to you as they are much easier to install and allow operators to stand outside the switchroom and any arc blast area so avoiding injuries. But you still have switchgear damage to repair / replace. CBS will make bespoke operators if one of their standard items doesn't match your needs.

I hope some of the above provides you with some ideas.

Kind regards

Donald Lane

Edited: 31 December 2011 at 10:54 AM by dlane
 31 December 2011 10:56 AM
User is offline View Users Profile Print this message



williamjohn

Posts: 178
Joined: 22 November 2010

Hi Donald
You comment on the high initial fault current from a generator. The design of the system MUST be based on the subtransient impedance of the generators. Otherwise the circuit breakers can blow up in normal service.

Practical problems are the other way. The fault current quickly decays and can fall below the full load value. Zero volts and full load zero power factor current may require more excitation than full volts and full load at system design power factor. Thus the protection must take this into account, eg by voltage restrained overcurrent protecton.

Best wishes
John
 31 December 2011 02:08 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Mr Chew,

Thanks for the information - some answers to your questions.

Some points to clarify.
1. Was the noise heard when breaker was being reset?
-Before - this is what promoted the elctrician to stand to the side wheh resetting, he sensed somethign was not correct.

2. Was the explosion caused when the breaker was being reset?
-Yes pushing downwards on the handle, although he claims to have barely touched it when it exploded (on both occasions). ABB confirm the contacts should NOT move when a tirp is being reset.

3. Was there any attempt to switch on the breaker?
-No - only on the second incident when the new breaker was installed and the electrician was trying to energise the circuit following a megger test, again he claims to have barely touched it when it exploded.

4. Was there any upstream tripping during the explosion?
-Yes on both occasions the 6.6kV transformer feeder and affected board incomer tripped. The 6.6kV is assumed to have tripped on instantaneous short circuit and the incommer for the switchboard on UV. The second point cannot be clarified due to the trip unit being damaged in the heat.


The second incident -

Some points to clarify.
1. Was there any upstream tripping during the explosion?
-Same as before

2. Was there any fault indication on the breaker?
-Nothing - it had just been installed. New out a box delivered form ABB. This was energising it for the first time.

3. What is the total generators capacity?
Approx 25MW but I will check for an exact figure

"We have also had reports that the transformer was recently tapped down due to excessive voltages burning out control cards in some circuits."

What was the value of excessive voltage and voltage noticed during the incidents?
The exsessive voltage was approx 720V - apparently when tapped down this went to 705. The switchboard downsrream is rated to 690V. The secondary of the transformer is 720V and not 690 as Iwould imagine it should be.

Please make a measurement on the phase - earth voltage.

We cannot power up this transformer yet - I will megger it today and check winding continuity to try and derrmime a fault.

"Other issues and small tell tales signs of arcing at the back of the downstream breaker (drive incomer) This shows some signs of shorting out or arcing but not enough to sustain the required fault current. We notice some scorching between phase and a small amount of metallic "spatter". Nothing excessive and discounted by the drive service Engineer who attended as insignificant. "

-If this is worth investigatring how would you suggest we do this? The arcing does not look to be significant and the ABB Engineer gives this the "ok".


What is the length of the cables?

-We are unsure - measuring them yesterday.

Thanks for taking the time to help and offer suggestions.

Mike
 31 December 2011 02:23 PM
User is offline View Users Profile Print this message



michaelfivey22

Posts: 27
Joined: 21 January 2004

Hi Dlane, Thanks for taking the time.

The main differnece between land based and applications on ships is that the ship has no neutral conductor, all circuit breaker are 3 pole only. The star point on the generator on LV (440 and 690 generator) is left flating with respect to the ships hull. On a HV ship (although not technically HV it is what we would call 6.6kV) the star point of the generator is tied to the hull using an impedance which will limit the fault current in a phase to earth fault. Some ships this detection of a full fault to earth does not even trip the generator breaker but provide the operator with an alarm.

Arcing faults are different to normal system faults in that the actual fault is of a higher impedance than a short circuit fault, consequently the arcing current of the fault is much lower than the short circuit fault current and the protective devices take much longer to operate. For a 690V system I would have expected initial direct contact but that would then be blown away by the arc as it is initiated and then sustains itself so the initial short may not be detected post event.

You say that your 6.6kV system is impedance earthed, is this also the case for the 690V system?

No - the impedence earthing is just on the 6.6kV generation level.

In my experience of these systems the impedance limits the current to a value below the arcing current so an arcing current cannot be sustained unless the impedance / resistor is faulty.

If the 690V system is impedance earthed then only a phase to phase fault can generate enough energy to sustain an arc so this is where I would concentrate my investigation. [/]

Agreed but we have been all over the system but have yet to find any resonable signs of a fault. you mention above;

"For a 690V system I would have expected initial direct contact but that would then be blown away by the arc as it is initiated and then sustains itself so the initial short may not be detected post event."

Can you please expand on this - the only part in the network we see damage is in the breaker - we can see no signs of damage i the swithboard bus bard apart from the small damage to the back of the downstream breaker cubicle.

A link based system may be more problematic than a tapchange switch, or is there a fault in one tap postion that is not present in others? . A turns ratio and insulation tests on the transformer may be prudent.

The link system looks straight forward and we are going to carry out the testing today.

Just a few words on safety. If the PPE has been specified correctly to match the arc energy levels then the wearer has a 50% chance of sustaining a second degree burn.

The 3 guys are actually doing well - all out of hospital and 2 of the three requiring skin grafts. They have all been very lucky. Worth noting though that during the investigation we realised that they were not wearing the full arc flash suit but only the visor and gloves with "bib" overalls. As you can imagin the buring was on the upper arms.

Thanks again for taking the time to reply. We have a lot of people on board looking at this and even people with a lot of experience in this area are getting a bit stupmed.

Thanks for the help


Mike
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.