Rotating Diode Open Circuit failure

[Acca]

We have been rectifying the above fault on machines in a marine application.

The machines are a SIEMENS 1FR6 454
1800 Rpm 450V 60 Hz 812 kVA @ 0.8 PF.
They are usually configured with at least 2 running insysnchronous at all times.

We suspect a contributing factor of the failure is the ships crew paralling out of synch with other machines onboard and/or with shore supply. Difficult to prove. They have automatic paralleling capability but regularly "practice" manually.

We have replaced the existing SEMIKRON semipack diode blocks with a EUPEC brand equivalent (as recommended by SIEMENS) which has a more robust housing. The SEMIKRON brand also contributed to the failure by cracking and internal contacts separting. Due to heat and rotating forces.

1. The open circuit failure of one rotating diode results in approx. 10% higher output voltage for the same excitation current. The exitation rotor is obviously a current transformer. Is this a common design as I have only witnessed a fall in voltage on other types of machines?

2. We are going to recommend installation of an In Synch relay in the Switchboard to stop out of synchronous paralleling. Where could I find more detailed information on these relays?

3. I have located some information on Rotating Diode fault detectors ( monitoring the stator excitation current ) on the internet. If the In synch relays are fitted is it usually common to have these fitted as well?

I would appreciate anyones additional comments or tips in the right direction. Especially if there is something else that I should consider.

ACCA

 

[boomsystems]

Put your hand on Alstom Make type SKE 11 synchrocheck relay. This is the most simple scheme to wire.

 

[busbar]

Depending the cost of a damaging synch attempt, there are protective products like Basler Exciter Diode Monitor {EDM200} that could be applied to shut down genset/disable excitation. That may inherently limit one more try at ‘getting away with’ manual out-of-synch breaker closing.

 

[busbar]

Based on genset “…approx. 10% higher output voltage,” a synch-check relay may provide only marginal protection in a failed-diode condition.

 

[Acca]

Thanks for the information to consider. Especially that the Synch relays may not be suitable. I was thinking that the synch relay would prevent the occurance of the diode failures.

The factory AVR Unregulated (Cct cards not connected) voltage droop is No load 495V/61.5Hz to 472V/58.5Hz at full load 1040A. The AVR regulated voltage droop is 458V/60.90Hz at No load to 442V/58.9Hz at full load 1040A/0.8PF.
After failure of a diode (O/C) this changes typically to 535V/61.5 Hz No load then increases to 550V/58.5Hz at 1040A. As stated previously the excitation current is the same.

The AVR will be regulated (cct cards connected)when the fault occurs which manages to bleed the voltage to 472V at no load. I did not apply load to see the response.

It has been a common fault on at least 1 generator for 7 vessels that are identical. With the oldest(1996)all 4 generators at some stage.

The cost of damage to the ships equipment I understand has been minimal. I have been told that they have only had O/C MOV's on power supplies. I think luck has been on their side.

As the fault occurs during paralleling the circulating currents are high.

 

[busbar]

I misunderstood the situation. I {apparently incorrectly} read that the sets were being paralleled after diode failure. My apologies.

If the synchronization attempts are causing diode failure, my suggestion is likely of no use.