Experiences with the setup of a micro-grid at TBEIC

Bill Whittenberger Uncategorized Leave a Comment

I have been engaged over the last 2 months with the setup of a small micro-grid at TBEIC, to be used for testing customer components that might be used in such grids. We have one customer who expects to test their intelligent breaker in a month or so using our setup.

The three primary components that define our micro-grid are:

– Powerhouse 25kW Load Bank, 480/3phase, load selectable in 5kW ranges, intended to represent a critical home or business power load that we do not wish to be disturbed when the utility grid fails.

– Battery Bank for storage, includes safety, disconnect, and metering functions. Generously donated to us by a local firm who wishes to remain anonymous. The bank is made from 12 volt lead-acid batteries, 40 amp-hour rating, 32 total, about 1000 lbs of lead. Nominal bank voltage is 384 volts. Nominal storage is around 15 kW-hours.

– Grid Tie Inverter, Princeton Power GTIB-480-100. This is an intelligent inverter rated at 100kW, meant to interface the 3 phase grid to DC storage, a PV array, and also a critical AC load. This 100kW box is very much over-rated for use with our 25kW devices. We use it because its what we have for now.

Other equipment to be used in the micro-grid are:

– Keysight PA2203A Power Analyzer, used for detailed characterization of 3 phase power

– NHR 9410-24 Grid Simulator. This is a 24kW device that will manufacture any sort of power needed. Most importantly in this application, it can simulate interesting types of grid failures such as brown-outs, phase drops, line spikes, harmonics, voltage notches, and the like.

– Keysight 34972A Data Acquisition System, used to capture data in real time.

– NHR 9200 Battery Tester, used to charge/discharge/evaluate the battery bank off-line.

All of the 3 phase AC equipment is fitted with what amounts to giant extension cords, which allows the equipment to be arranged in any electrical sequence. In the simplest arrangement, the inverter is plugged into and powered from the wall disconnect, and the load bank is plugged into the critical load connection on the inverter. And the battery bank is connected using a DC connector to the inverter as well.

In a more complex connection arrangement, the following will be plugged together in order to power the inverter – wall disconnect, grid simulator, customer breaker, power analyzer, inverter.

Commissioning of this micro-grid arrangement has provided a numbers of surprises and challenges that allow a good engineer to make a living:

Powerhouse load bank – This was initially plugged right into the 480 wall disconnect to test its function. It also has a 110V cord that powers its logic and instruments that needs to be plugged in. As is happened, a wall plug with a GFCI was conveniently located. When both cords were plugged in (with nothing turned on) the GFCI tripped. The unit functioned as intended when plugged into a conventional 110 outlet. Hoping to avoid electrocution, I had some dialog with Powerhouse, which resulted in connecting the ground wire in the 3 phase cord to the ground stud on the box instead of the plug provided. Now everything plugs in without tripping the GFCI. But when you turn on the load bank, it still trips. Hmm. I decided to keep it plugged into the conventional 110 outlet and moved on, knowing that the case was grounded and that I would be safe.

Grid simulator – Initially used this to power the load bank. Set up the control to provide 480 3 phase, turn it on. Turn on the load bank, it would draw power for a half second, then not. Hmm. Turns out that I needed low-side connections for each phase on the back of the simulator. Read the directions next time! After fixing this connection, the simulator has quickly and easily done everything I have asked of it.

Power analyzer – next step splice this into the power supply chain to the load bank, and learn how to use it. This is a really nice unit, lots of buttons, touch screen, storage, graphics, and the like. After pushing buttons for an hour, the unit locked up. Very pretty screen, but no response to any of my inputs, including the USB from my computer. Off to Malaysia for repair, after extreme fun getting the connectors loose, and discovering that somebody has re-purposed my box. Grr. I hope to have this back in mid-Feb.

Battery bank – This donated unit came with no instructions, so it had to be figured out. There were wires inside that powered the instruments that had been disconnected to avoid discharging the batteries. The box came to life nicely after I found and connected those wires. After locating and purchasing the correct (and expensive) DC connector I was able to connect to the NHR charger and run several charge/discharge cycles. Working well, except that it feels like the actual capacity is low as compared to the ratings. Need to examine the data carefully.

Connecting battery bank to inverter – First connection, turn on the bank (inverter is still off) pow blow the 30A fuses on the battery bank. The inverter has internal capacitors that charge when the battery is connected. Princeton warns you that you might need to implement a soft-start circuit. Not wanting to mess with this, I tried 50A slow-blow fuses, which is still within the capability of the components and cabling. Problem solved.

Inverter startup – This is the interesting part. Needed to fix the phase rotation on the grid connection. Tens of parameters to set up using a clunky screen with a dial and click input. After some doing, I got the web interface to my laptop up and running, which makes setup MUCH easier. Managed to get the inverter in run mode. Interface to load bank works correctly. But not drawing power to/from the battery bank as expected. Why? There is considerable discussion in the manual about needing an isolation transformer if your DC device is tied to ground in any way. Mine is not, so no transformer required. The manual does not mention that if no transformer is used, you need to add power jumpers inside the cabinet. After tracing wires in the cabinet trying to figure out why there is seemingly no connection between AC and DC, I stumble upon a small sticker that mentions jumpers needed if no transformer. They might have mentioned that in the manual! Jumpers installed. Now the inverter trips out claiming low battery voltage whenever I try to go to run mode. Fault buffers show ridiculously low battery voltage – in the 200s. Princeton believes that this is startup transient problem created by the battery bank being too small. They have a software/hardware fix but it’s expensive. Trying now to figure out how to dodge this problem without a big spend.

Stay tuned ….