6 SCR Heater Controller Feeding Transformer Coupled Load 1

[Laplacian]

What is the best arrangement for a single zone of multiple parallel heater probes from a load balance perspective on the SCR controller? The application is as follows:

* 6-SCR,3-phase,480V controller, phase angle fired
* 112.5KVA, 480V Delta - 100V Wye (58V ground), 3-phase encapsulated transformer
* 108 - 750W heater probes (can be any voltage)

I was considering connecting them all in parallel delta for 90V rated probes.

The issue is potential half cycle faults due to load imbalance when individual probes fail and open up. Probe replacement is difficult and requires costly downtime.

What parallel connection would result in the most balanced load as seen by the SCR controller through the delta primary transformer winding?

Would it be a 4-wire wye connected heater arrangement? Ungrounded wye? Or delta? Note that each probe is individually protected. I can specify 50V or 90V probes from the manufacturer.

Thanks for any input.

 

[Skogsgurra]

I do not quite understand why you would get half-cycle problems if one of the "probes" opens up? The probes are AC fed? Aren't they? If one of them opens, it just means that current is reduced. Both plus and minus half-waves. So, all you will get is a (slightly) unbalanced load. But no half-cycle (transformer core DC saturation) issues.

The real risk is that you get half-cycle problems due to inadequate triggering. Or problems with turn-off when running low heat (phase-angle large).

DO NOT USE UNGROUNDED WYE! I.e. if ungrounded wye means no return wire to transformer star point. It is a physical possibility - but tricky. Parallel delta or grounded wye? I would say it doesn't matter. Wye will make measurements a lot easier. But you need heavier gauge cables. It is a matter of taste. If distance between controller and "probes" is short - then I would select grounded wye.


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Laplacian]

The controller manufacturer stated that a ~10% load imbalance would result in half cycle errors. The probes don't really last long in this application (1800F).

The original voltage was too high, requiring a very light gauge heating element. The high ground voltage also resulted in a lot of ground leakage current (arcing and sparking within the probe) leading to failure. The manufacturer recommended to operate the probes at as low a voltage as possible. It is much more economical to put the transformer on the controller output.

If there is no technical advantage between grounded wye and delta, we will go with delta. The transformer X0 bushing will be solidly grounded.

What are the issues you've experienced with phase angle control mode? You state inadequate triggering and commutation issues. Are there any mitigation steps that can reduce these scenarios?

Thanks again.

 

[Skogsgurra]

Yes, inadequate triggering (can occur if your transformer has a high leakage inductance so that thyristor holding current isn't reached before gate pulse goes away) can be mitigated by connecting a snubber (microfarad and 100 ohms) or an incandescent lamp across each transformer primary.

Cannot understand why load imbalance should be a problem. Are you sure there isn't a misunderstanding somewhere?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Laplacian]

A asked twice. All the information I read indicated that phase-angle firing for a transformer coupled load was the best solution for unbalanced loads. I don't know why the manufacturer's application engineer said otherwise.

About the snubber: Connect a 1uf capacitor and 100ohm resistor across the primary terminals; phase-phase correct? Capacitor and resistor in series or parallel? I appreciate your help.

 

[Skogsgurra]

Series connected. Resistor reduces current (you need more than 1/4 W) and capacitor reduces power in resistor. You need this if the gating pulses are short. If you have long, chopped trigger pulses it usually isn't a problem.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...