Harmonic Filter - Cost, Size?

Only 7 to 8% THD gives your generators problems? What kind of problems? How can you be sure that by filtering you will clear up the problem? Maybe all you need is a set of line reactors between the generators.... And, how much overall current are your talking about?

Hi, thx for the comment.

We dont have any problem with the generators, i might have been a bit unclear with the problem :)

The problem is that the refrigerator containers that we feed power for driving the compressor, switches off from electric power and starts up on diesel instead when THD goes above about 5.5 %. The THD vaule goes up as we increases the load on the propulsion motor.

/Lars

Some more explanation. My idea would be to put in detuned filters on the downside of the transformer that feed the containers.

The maximum current to the containers is 800A.

So the problem is that the refrigerated containers are supposed to use your power but instead they are starting their own diesel powered cooling units?

And the container coolers are sensative to the THD? Thus refusing to allow external power to be used?

yes, that is our problem.

Thx for your fine answer but the problem with the containers, which I should have been clearer about, is that they are on trucks and they are not ours as we just transport them as I work on a ferry. So for us to do any kind of modification on the containers is not doable.

Is the idea about putting in detuned band pass filter on the downside of the 400V transformer not so good?

Connecting Filter downstream of Auxiliary transformer which is feeding in to Freezer Container will not help.

Reason being that the distortion is in 6.6kV output voltage generated by 3 phase synchronous alternator due to Harmonics Current fed from inverter. This is further being reflected in the 400V side of Transformer. The filter removes the harmonic currents. Hence any filter has to be at 6.6kV bus.

The picture of connection of Freezer Containers and Diesel Generator is not clear to me. What I understood earlier that each container has AC electric motor driven compressor + condenser unit. Individual freezer is normally supplied from common ship power and in case of problem with ship power there is a 400V, 3 Ph DG Set with AMF and changeover on the deck, common to all containers. The AMF relay of DG is malfunctioning due to harmonics. Please clarify connection between containers and Ship.

The ships power comes from four diesel generators with a voltage of 6.6 kV. This 6.6 kV is used the propulsion converters. It is also transformed down to 400V for different three phase loads and among them a 400v/400V transformer that feeds all the power outlets for the freezer containers. The reason for the second transformer is to separate possible earth faults from the ships system.

There is no common DG set that will be switched over to incase of problem with the ships normal power. Every freezer container/truck can switch between using its own diesel set or an electric powered set. This switching is automatically.

Noted that there are two Distribution Transformers between DG sets and Container Freezer units. First 6.6kV: 400V for main ship bus and second 400:400 isolating transformer between 400V Bus of Ship and Freezer units. The isolating transformer is probably used to convert ship floating neutral to grounded neutral for freezer units.

When designing core of Transformer, for economy, the designer selects operating flux density of at slight saturation 10% to 15% above linear range of magnetising curve. Because of this saturation there is slight harmonics in secondary winding. Which has no problem in case of normal supply. In your case the supply is already distorted and possibly the harmonic contents at 400V supply to freezer bus shall get worsened due to saturation in the core.

We cannot do anything to 6.6kV to 400V Transformer. Hence the harmonics at this bus is accepted.

We can definitely do something to 400:400V isolating transformer so that any distortion by virtue of this is eliminated. So as a first step, disconnect supply to containers from Secondary of Isolating transformer and connect directly to primary input of isolating Transformer. Try out Freezer containers at different load of inverter - if no problem is experienced order special isolating transformer and replace existing with new. The replacement transformer will have relatively higher cross section area of core and of material of better quality.

If above trial does not work - means the harmonics at ship 400V are of sufficient magnitude and required to be filtered. In this case suggest to install Harmonic filters at 6.6kV Bus.

Keep me posted of progress.

Best of luck.

This problem can be mitigated with a passive harmonics filter but its design will be difficult if not impossible as the impedance of the source is not so low and changes drastically with the number of generators operating. You will need to use many smaller banks in parallel to adapt to the rapidly changing load. A fast controller (single cycle) with its corresponding fast semiconductor switches will be needed. The load also varies drastically when you are in a storm. This makes the phase and the amplitude of the harmonics change every time you hit a wave. The reefers load variation is probably negligible in this setup but this type of load is known to be highly variable as they turn on and off almost randomly.

In these applications, you are better off with an active harmonics filter that will not be susceptible to the resonance problems of the passive alternative. The cost will be higher though.

How about using a set of 1:1 isolation transformers? The harmonics would have a hard time penetrating due to the iron core of the isolation transformer. You can buy phase shifted windings such that they actually eliminate certain harmonics. What I mean by phase shifted windings is a technique where a number of turns from the adjacent transformer are placed on the first transformer and so on.

Most lower harmonics (up to 30-40) pass through the standard isolation transformers with little attenuation. The main exception are the multiple of 3 that are blocked.

This will not solve the OP problems with 11 and 13.

The OP could use a specially designed mitigation transformer to attenuate the 11 and 13. This transformer could be used only on the 415V feeder to minimize its size. The net can list plenty of suppliers.