HomeMy WebLinkAboutTrip Report to UniEnergy Technologies 7_23_14ENERGY EFFICIENCY EVALUATIONS
Site Visit to UniEnergy
Technologies Battery
Company
Prepared for Kotzebue Electric Association
Dennis Witmer
7/23/2014
Trip report from visit to UET at their headquarters in Mukilteo, Wa. Several utility sized Vanadium flow
batteries were observed in operation, company appears to be close to shipping initial commercial
product for demonstration. Technology appears to be more developed than other flow batteries
evaluated in KEA RFI process. Cost and performance should be modeled.
Arrived at UET for meeting.
UET is housed in a 69,000 square foot building that they own, though currently part of the building is
occupied by another company. They will be taking over the entire building at the end of the year. The
entrance to their offices was not obvious from the street (the other business is currently occupying the
main entrance to the building), and the door was locked when we arrived. It was pouring rain, I tried
another door, found it unlocked, and was able to enter their manufacturing space, a large room with no
one present. We then went back to the main door, and met Charlie Vartanian, the Marketing director.
We signed into the visitor log on an iPad. Outside reception area is office space for approximately 20
people.
Met and had a brief discussion with COO Rick Winter and consultant Garth Corey, author of the white
paper previously reviewed.
Meeting began in conference room A, with introductions from:
Rick Winter, COO
Liyu Li, PhD, Chief Technical Officer, Co-Founder
Charles Varanian, Marketing director
Russ Weed, VP Business Development and General Counsel
Ted Volberding, Director of Manufacturing
Kathyrn Oseen-Senda, PhD, Senior Systems Engineer
Dave Ridley, Director of Electrical Engineering
Jon Horner, Mechanical Engineer
I introduced myself, and briefly discussed the previous battery project in Kotzebue. There was
considerable interest in what happened with Premium Power. UET discussed the environment of a few
years ago, when Premium Power would “stand up in meetings and make great claims about their
battery that were difficult to either prove or refute”, and that they “poisoned the well” by creating
unrealistic expectations about both price and performance of battery systems.
UET is familiar with diesel economics based on their discussions with Hawaii and PV arrays. However,
there was much discussion about the total cost per kW-hr of energy vs fuel only costs—and the difficulty
in even getting real fuel costs in remote villages.
There was a discussion about arctic climate issues. The electrolyte currently in use does not freeze at
40 below, but it does become viscous and less conductive. However, it is not clear that the power
electronics is hardened to the same temperature regime.
Charlie Vartanian had a copy of the KEA Jessie Logan paper submitted in the summer of 2012, although
his copy differed from the report available on the DOE web site, in that it had a picture of the Premium
Power flow battery on it.
The initial introductory meeting was followed by a “factory tour”. I requested permission to make
photographs, and was told that I could take some photographs, but that they were cautious about which
things I could make photographs of. I requested permission for each photograph I made from Rick
Winter.
First, the factory was clean and well lit, and appears to be in a “scale up” mode. The intention is to use
this factory to assemble the containers that form the battery system, with 4 DC battery modules for
each inverter module, each installed in a custom built 20’ conex box. Each DC battery module holds
large electrolyte tanks that occupy about 80% of the interior volume of the conex box, plus 3 stacks,
each of about 50 cells of 4000 cm2. One battery box produces about 250 Volts and 1000 Amps DC.
There were two demonstration units operating inside the factory, the first nicknamed “ELVIS”
(complete with a large poster behind it) for “Electric Leveling System” and Angus (named for the lead
guitarist for AC/DC) which is cycling from full charge to full discharge every 30 seconds for 2 months
(another month of testing planned) with no observed degradation in performance. ELVIS was the first
commercial scale system built, and has been operating for 6 months.
The battery system is comprised of several battery units (usually 4) and a power electronics unit (housed
in a separate container). Each container is a custom version of a standard 20 foot conex box, built with
heavier steel, and one foot higher than the standard box (dimensions of 20’X8’X9.5’) and designed to
hold 40 tons of weight (well above the 48,000 pounds of a standard container). The conex box has a
welded bulkhead in front of the electrolyte tanks designed to function as a secondary containment for
the entire electrolyte volume, eliminating the need for a secondary containment on site. The system
also has an “anti syphon” design to prevent the transfer of electrolyte to the area in front of the
bulkhead in case of pump failure. There is a lower bulkhead in the stack area designed to hold 2X the
volume of electrolyte normally contained in the stacks.
Each battery compartment has two electrolyte tanks, one about twice the size of the others. Pumps for
the electrolyte appear to be submerged inside the tanks. The tanks are made with a special process,
where a full sized mold is heated and rotated, then loaded with plastic pellets which melt and coat the
walls of the mold evenly, creating a thick, seamless tank. If a larger energy capacity battery is desired,
larger tanks can be manufactured, which can be placed inside larger shipping containers without
affecting the construction of the stack area.
Each battery container has 3 stacks, all fed from a common manifold, so all stacks and cells are exposed
to the same electrolyte composition at any point in time. A single cell at open voltage is used to
constantly measure the state of charge of the system. The cell area is approximately 4000 cm2. Stacks
are built in China by UET’s partner. The manifolding for the stacks was covered with plastic shields, and
apparently was one of the sensitive items UET did not want photographed. Each battery container can
provide 150 kW of power, and 600 kW-hours of stored energy.
There was also a controls package inside the stack area, enclosed in a separate metal box with an air
intake from outside that is vented into the stack area. Apparently there is some kind of contamination
that comes off the stacks that can coat surfaces that is electrically conductive, and can affect the
performance of the control circuits. Control of the entire battery string is from a single master
controller, for speed. It was not completely clear how quickly the system can respond—there was
mention of 8 millisecond response of the battery, but between the control system and the inverters, the
total response time is something like 100 milliseconds. The control system is a PLC design supplied by
Siemens.
A “string” of 4 battery modules and one power conversion unit can deliver 600 kW AC, and store about
2.4 MW-hr of energy. This is intended to be the basic unit that UET will sell. Based on the phone
conversation of 7/15/14, current pricing is estimated at $1000 per kW-hr of storage, or $2.4M per string,
of which 80% is battery cost, and 20% is installation cost. Note that the cost of a single string is above
the currently available funds in Kotzebue, and that installation costs (including shipping) are likely to be
higher in Kotzebue. Prices are expected to drop by about 25% in the 2015-2016 time frame.
The inverters are large industrial inverters—Packer and AEG have been the inverter suppliers they have
used to date. Angus was operating with a Packer inverter.
The factory is designed so that the conex containers never come inside the factory area—only the metal
rack that supports the stacks is moved around the assembly area. The conex boxes (custom built in
China for about $9,000) remain outside, and have the tanks installed in them. The stacks, mainfolds,
electrical system, and control system are assembled as a unit and installed into the waiting conex box.
In the production mode, the assembled system will have electrical and pressure tests done at the
factory, but there will be no electrolyte loaded into the system until it is at the site where it will be
operated. This makes the unit much lighter to ship (10,000 pounds vs the 80,000 pounds once the
electrolyte is loaded), as well as less hazardous. Electrolyte will be shipped directly to the final
installation site. (Note for Kotzebue: Total shipping weight is very high—about 350,000 pounds for a
full string of batteries, electrolyte, and power package.)
Units come with a one year warrantee and a 20 year service agreement (not disclosed what those fees
would be). At the end of 20 years, UET will decommission and remove the unit at no cost (paid for by
the value of the vanadium that can be recovered).
The unit is designed to minimize the amount of site prep that needs to be done—obviously, the unit is
very heavy, so a solid concrete pad needs to be constructed. All connections between conex units is in a
single penetration located at the top of the box just behind the large doors. The conex boxes can also
be bolted together and bolted to the concrete pad for earthquake mitigation.
The design philosophy is to minimize the work required for field installation, and to do all the critical
work at the factory, where it can be done by knowledgeable skilled workers under careful quality control
conditions. The units are also designed to minimize the risk of damage in shipping. In some ways, the
design philosophy is very similar to the new small scale nuclear philosophy—build units in a factory, ship
them to the site, load the chemical, and offer decommissioning as part of the package.
Outside the factory on a large concrete pad near the shipping dock was a “string” of 4 batteries plus the
power electronics box, currently being operated (know as “Johnny Cash” for obvious reasons.) In
addition, a set of five empty conex boxes were on site, the containers for the next string to be
assembled.
There did not appear to be any units currently being assembled, other than a single manifold assembly
of curled tubes which I was told not to photograph. There was no “boneyard” of old stacks. We walked
past the R&D areas, but did not go in them. The factory felt pretty empty, but it seemed that that was
intentional, to allow for the planed future assembly of systems.
Current production is “zero” (said with a laugh), but plans are to eventually produce one battery
container per week, one “string” a month by the end of 2014, ramping up production to one string a
week by the fall of 2015. In about 2 years, it is expected that the current factory will be at capacity, with
the next manufacturing site to be chosen “next to a dock or on a rail line” to aid with the shipping of the
heavy units.
Note was made that while vanadium does not have a high power density per unit area of membrane, or
by volume, the energy stored per acre is very high—about 40 MW per acre, or 160 MW-hours per acre.
Figure 1 UET data display in front, electrolyte tanks for 20' conex box unit in back.
Figure 2. View of "Angus", showing three cell stacks (white), covered manifolds (center) and electrical control box (on left).
Note white secondary containment bulkhead . Unit was in test, switching from full charge to full discharge every 30 seconds.
Figure 3. "Johnny Cash" unit behind factory, currently operating in full cycle charge/discharge mode. Note radiators for heat
removal on top, as well as vents in doors of electrical unit in foreground.
After the factory tour, the meeting reconvened in the conference room. Most of the remaining
discussion was about the possible deployment of a battery of this type in Kotzebue. First, it was noted
that the Premium Power battery was not ideally sized for Kotzebue, as it did not provide sufficient
power to cover either diesels off operation (not likely anyway, given the level of penetration) or even
sufficient safety margin for an EWT turbine trip. After some discussion, it was agreed that a good size
for Kotzebue’s current configuration would be a 2 string, 1.2 MW, 4.8 MW-hour battery, though the cost
of such an installation is beyond currently available funding.
There was also some discussion about the proper way to install these batteries, especially given the
experience with the Premium Power battery, which was (per agreement with the supplier) simply
parked outside. Given that this would be the first of a kind in the arctic, it was felt that some kind of
protection from the drifting snow was required—probably a steel building of some kind. While this
construction would add to the installation costs, it would provide some measure of protection for the
operation of the unit (the issue of snow blowing into the electronics being one of the major
considerations). One issue raised by UET was the need for some kind of external vent for each of the
battery units—apparently there is some material which should be vented to the atmosphere and not
inside a building.
There was also some discussion about data needs for modeling—UET was very concerned about
releasing electrochemical data—they have not shared that data with anyone to date. They openly
discuss their overall efficiency (65-70%), the fact that there are limits to the charge/discharge rates at
the ends of cycles, and the level of electrical parasitics (8 kW for all pumps on a string of 4), they were
quite reluctant to discuss. They suggested a non-disclosure agreement with KEA before they release any
data about the specifics of their performance.
General Company Info:
Company is wholly owned by a “private investor” (not a venture capitalist) who has invested on
the order of $500M to purchase the entire supply chain for these batteries, all the way from
vanadium mines, stack developer, and UET. IPO may happen, but “six years away”
Stack and electrolyte both come from China. Stack developer has been building stacks for 15
years.
Management is old and gray, engineers are young and seem happy. Seems like a good group of
people
There is a reasonable level of IP protection concerns.
General Impressions of Battery:
UET has built several demo units for “inside” evaluation, three of which are currently operating
Battery appears to be competently designed, with attention paid to leak prevention, secondary
containment, transportation and shipping, and minimization of electrical parasitics
Control system and inverters are from known reliable suppliers
Turn around efficiency of 65-70% and cost of battery both will limit the value of this battery
compared to heat sales
Risk of “First of a Kind” deployment in Alaska.