VFD power regeneration

[thinker]

Our customer initially ordered a VFD based test stand. Electrical dynamometer motors operate only in power absorbing mode. The 4-quadrant VFD (with PWM IGBT based front end) was expected to regenerate the shaft power (about 3 MW) back to the grid. The problem occurred when the customer failed to receive from utility a permit for co-generation of power. Now under consideration is a possibility to dissipate that energy via large load bank. One of suggestions was to create a large DB resistor system (like in non-regenerative drive) tapping the DC bus. Other suggestion (by drive MFG) is to use existing front end converter for regeneration of power into AC load bank. Question to community: is this something for real? Common understanding is that one needs front end converter in regen mode synchronized by the grid voltage and frequency. Is this possible to route the shaft power through front end converter into AC load bank ? Reference to any published materials on the subject would be greatly appreciated.

 

[jraef]

Wow, sad that this was not vetted properly before the drive was purchased. Do you know for sure that your other facility loads cannot absorb that load? You only need to send it back to the utility if the regen load exceeds the facility load.

Typically an AC load bank is just a bunch of resistors burning power off as heat. If you have a boiler for any other reason, use that to capture that energy. Otherwise, you would need a system that can control the amount of load added so that it exactly matches the regen energy, otherwise if you have less regen than the load you connect, you will pay to feed the load bank from Utility power. Typical AC load banks are not that sophisticated because the common use it just to exercise standby generators. You will have to contact someone with experience in doing this right.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[waross]

What are you testing? Electric motors loads plus losses will be greater than the regenerated power.
If you are testing diesel engines it may be well to start by re-evaluating the type of dyno loading. If you can't re-gen into the grid, a Motor-VFD combination may not be the best solution.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

Yes, what are you testing?

 

[thinker]

These are wheel dynamometers of a large vehicle. The prime mover is a vehicle's engine. Dynamometers are for loading only. On each side (left and right) there are two dynamometer motors connected to input shafts of summing gearbox. Output (high torque) shafts of gearboxes are connected to test specimen. Yes, it was surprise for us that proper vetting process was not provided. Customer still hopes to get permit for power regeneration to the grid, so the 4-quadrant capacity of VFD must be preserved. They want to have an interim solution for testing by dissipating absorbed energy via load bank.

 

[waross]

As a suggested avenue of investigation, have you considered liquid rheostats?
I have used brew your own load banks comprised of a plastic barrel filled with salt water, with a couple of truck leaf springs for electrodes. This was long long ago and far far away and not subject to approvals or safety officers.
The point is, a manufactured liquid rheostat will have approvals and will be somewhat controllable.
Hopefully you could trim the rheostat to stay just below the power export level.
Is anyone familiar with the capacities available with liquid rheostats, and current manufacturers?


Anecdote
This thread reminds me of an installation that never happened for a co-generation scheme.
A small lumber mill wanted to save power with a steam turbine generator using the boiler that supplied the dry kilns.
The supplier may have been an expert in power generation but did not know crap about drying lumber.
The proposal was to drop the steam pressure with a non-condensing turbine and use the turbine exhaust to supply the kilns with low pressure steam.
For a given capacity a non-condensing turbine is much larger and more expensive than a condensing turbine.
The volume of low pressure steam would have required quite large and expensive piping.
I met the designer on site. He told me that this was the only way to go.
He kept repeating the mantra;
"You can't dry lumber with high pressure steam!"
This while standing in front of four dry kilns drying lumber with high pressure steam.
The kilns would not be able to efficiently dry lumber with low pressure steam applied to the high pressure steam coils
Power production would be limited by the steam through put of the kilns..
There were a couple of other design glitches due to unfamiliarity with lumber kiln operation.
And the big one._
The power authority was never contacted and would not have issued permits to allow co-generation.
My solution:
Leave the kiln system untouched.
Install a condensing turbine and share the steam with the kilns with priority going to the kilns.
Keep power production below the mills usage to avoid feeding the grid.
Fortunately, the parent company decided (for tax reasons) to shut down the mill and shift lumber production to another country.
Disaster avoided.
Anecdote off.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jraef]

The issue with regenerating onto the incoming utility line, if you are in the US, is that to do so, you must have a "Net Metering Contract" with your power utility, the one that allows you to spin the meter backward. But in order for them to give that to you, they will need to see that the generating equipment (in this case the VFD) has a safety approval for "Line Interactive Equipment", typified by having a UL-1741 approval. This is what ALL on-line solar inverters must have for example. The reason is, they must have the ability to detect and shut down if the line side is off, such as in a power outage, and this feature CANNOT be defeatable. Why? Because if they don't have that, and a utility lineman is working to fix the line in an outage, your regenerative drive could come on and kill him!

Currently there are NO VFDs on the market that I am aware of have UL-1741 approval as yet. It is VERY difficult (translate: expensive) to attain and there is currently no market for them large enough to justify the expense. So their hope of someday getting this ability is more of a pipe dream. You might want to let them know.

But again, if they can use the energy INSIDE of their own service, i.e. run other loads that are connected, then no problem. The PoCo may however insist on an automatic line disconnect that opens in the even of a line failure (same issue). Some AFE drive include that, some do not.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[davidbeach]

I think I saw a UL-1741 dyno package several years, but can't readily find that application at this point. Ideally the utility should know of all "sources" on the system even if the source always has enough load with it to maintain an import condition. Small inverters don't individually make much difference, but enough of them can make a significant source of fault current.

 

[waross]

An idea and a question Jeff.
I assume that a VFD has two internal connections to the incoming line.
One connection for power in and out.
A second connection to the circuit board which holds the logic and does the timing and firing.
Is it possible to separate these connections so that the regenerative current will be synchronized with the grid but instead of being connected to the grid, be connected to a load bank?
Alternately a high impedance connection between the control circuit and the power circuit. The control circuit would be connected directly to the grid.
The power circuit would be connected through the high impedance, but would have a load bank directly connected to the power circuit.
Only a small amount of power would flow from the grid to the load bank under light loading and only a small amount of power would flow back to the supply side when regenerating.
The VFD manufacturer may be the best to answer this.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[FreddyNurk]

I was of the understanding through dealing with a US based inverter manufacturer that they had finally achieved UL 1741 compliance, although I cannot remember if they actually did, and at any rate I wasn't concerned as it wasn't a relevant standard for my location.

I had thought that the workaround in the absence of UL-1741 was an extra protection relay capable of detecting islanding events and shutting the inverter down. Certainly that has been the alternative here in Australia when compliance with grid connection standards hasn't been granted for the inverter (i.e. AS4777, noting that as I recall, it has size limits on applicability of inverters).

 

[waross]

Another suggestion that the VFD manufacturer may be able to accommodate.
A VFD configured with two BACK ENDS.
When the motors on one back end are regenerating, (all the time in this application), the other back end could be powering a load bank instead of a motor.
The power would have no connection to the grid.
This could be easy or difficult. If the control scheme may easily be reconfigured, easy. If a new design and control circuits and algorithms are needed, difficult.
It depends.
Remember, this is a 3 MW VFD. A solution that may be prohibitively expensive for a 3kW VFD may be relatively economical for a 3MW VFD.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

UL-1741 just means the solar inverter can be directly line connected without any other relaying to protect against islanded operation. It's a way to combine all the functionality into a single device so the customer doesn't have to spend as much on protective relays and associated equipment as they spent on the inverter.

If you wanted to be argumentative, using a setup where the regenerated power was also consumed by the customer would negate the need for even considering UL-1741 as a requirement, yet the site would still have an inverter which ~could~ become a generator. We have a line regenerating inverter here and it's not UL1741 rated.

Utility workers who don't consider every line they work on as being live don't have good odds for making it to retirement...

 

[jraef]

No argument from me on the reality of having a Regen drive in the facility being essentially the same, I'm just pointing out that if you want to feed BACK to the utility as a primary energy sink, they don't allow it without that Net Metering Contract, for which they will require UL 1741. I've been going through this for a couple of years now with an alternative energy storage company wanting to use Regen drives to put mechanically stored energy into the grid. Technically, it's not a problem doing it. Legally, it's a quagmire.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[waross]

Hi Jeff.
Thank you for your kind reply.
The issues surrounding UL 1741 are interesting and are valuable for future reference.
They may be moot here.
The customer wants to sink power from a dyno test.
If they cannot back feed into the utility than UL 1741 becomes moot in this instance.
They have to sink the power somewhere. If the grid is not presently available some other energy sink must be provided.
If this is a small testing lab, with a low base load, then it may not be worth the effort to try to supply the plant load.
The first issue is to use or dissipate the energy developed by the dyno testing.
The second issue is to keep the option open to someday regenerate into the grid, even if that never happens.
So the problem may become:
How to use a VFD to provide some type of load bank to dissipate the energy developed by dyno testing while keeping the option open to someday back-feed into the grid.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[thinker]

Gentlemen, thanks for your inputs, very informative and useful. The very last sentence in Bill's posting actually repeats my OP. Yes, we will try to keep the whole 3 MW regen system intact, and add to it a large energy sink. We thought, the best way would be to add DC resistive load bank, connect it to DC bus, and make the system operate as non-regen with DB resistor.

 

[LionelHutz]

First you would need to ensure that you could actually draw power from the DC bus. The typical way this is done is via a chopper IGBT with the duty cycle controlled by the bus voltage. The problem is that a regen VFD actually attempts to control the bus voltage compared to a 6-pulse diode front-end where the bus voltage simply rises when the VFD goes into regen.

So, switch to an AC load bank. Well, you would have to switch on enough load to keep the VFD from supplying any power to the utility. But, too much load means you are using grid power to feed any difference and paying for that electricity. So, you'd want a bunch of reasonably sized steps and a controller to continually step them in and out as the VFD output changes.

I highly doubt any kind of trick could be played to make the VFD regen AC into a big 3-phase fixed resistor bank.

 

[FreddyNurk]

As a matter of fact, now that I think about it, there was an inverter I came across a few years ago where a load bank (well, actually it was a boiler, but the principle in this case is the same) was connected to the DC side of an inverter (sometimes known as active front end). There's probably enough differences to make it quite a pain to duplicate, but its possible.

In that case the inverter was quite capable of pushing back into the network, but couldn't, as there was no energy source on the other side of the DC bus, just a load. It wasn't connected to the network either so was a moot point there too.

As for the utility issue, its generally their network, their rules unfortunately. Easiest thing to do is often put a relay in, commission it, and keep the utility happy.

 

[waross]

Easiest thing to do is often put a relay in, commission it, and keep the utility happy.
So everything is working great we have a large vehicle on test.
We ramp up the power output of the diesel engine.
The regen energy exceeds the power draw of the plant and the relay trips it off line to satisfy the utility.
Everyone is happy. It worked just as intended.
Well, almost everyone. The poor shmuck who paid a lot of money to build a test stand to test large vehicles and has just found that his test has been truncated may be a little downcast.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[FreddyNurk]

waross, you're correct, although I admit I'd forgotten about the original requirement for using the equipment as a test stand. Having worked for a utility for a few years, I am aware of just how hard it can be to convince them to accept alternatives to their arrangements.