 |
 |
Electrical Distribution Technology:
What it Takes to Make it Go
Click here for a printer friendly version
of this article (56k)
Robert Broadwater could be called "the reluctant entrepreneur."
A computer engineering professor, he is currently also serving
as president of Electrical Distribution Design, Inc., a firm
that helps the Electric Power Research Institute (EPRI) market
and support software for electric utilities. The firm has a sound
fiscal base, has six full-time employees and several part-time
employees, and provides support for two Virginia Tech graduate
students.
However, Broadwater would like to replace himself with another
president, and did not want to be actively involved in the business
activities when the firm was established.
In 1995, a Virginia Tech research
team led by Broadwater developed workstation software that could
save a medium-sized electric utility $1 million a year. The DEWorkstation
software helps improve the design and analysis of large electric
distribution systems. The software takes data, such as load research
data, circuit measurements, and customer data that are stored
throughout the utility and relates it to a circuit model. Fifteen
different analyses applications can then be used to analyze the
utility's circuit model. DEWorkstation helps improve reliability,
reduces cost, makes better use of engineers' time, helps engineers
solve problems they previously couldn't analyze, and serves as
an error filter for utility data. The workstation is used by
utilities in the United States and in foreign countries, including
Chile and Thailand.
EPRI licensed technology from Virginia Tech for its member utilities,
and EDD was established to support the software. EDD is not a
commercial outfit, Broadwater said. "We operate more as
a support arm for EPRI." EPRI provides the DEWorkstation
to its member utilities at no cost. Getting it operational, however,
costs between $500,000 and $1 million. "It's a significant
investment for a utility, even though it typically pays for itself
in a year," Broadwater said. The utilities hire EDD for
training, user support, software maintenance, customized features,
and large-scale implementation.
Although it was initially ordered by more than 100 electric utilities
across the United States, the DEWorkstation has been implemented
in only a very small fraction of that number. "One of the
critical issues was that EPRI did not provide for long-term support,"
Broadwater said. "Members were thus concerned about what
would happen if they left the organization." Another issue
was that during the first few years most EPRI members were not
even aware that EDD was available for support. Even today, many
EPRI members do not know of EDD. Part of the early lack of recognition
was due to the software being shipped for EPRI by a third party.
Getting a sufficient installed client base was not the only problem
for EDD. EDD suffered initially from its own management challenges.
"Initially I was not interested in starting up the business,"
Broadwater said. "As a faculty member, I did not have the
time to be involved." However, after encouragement from
the university administration and start-up company specialists
at the Corporate Research Center, he agreed to serve as a technical
resource. "Unfortunately, after approximately 18 months
of operation, I was forced to take charge of business matters."
"We started EDD on a shoestring," Broadwater said.
He shares ownership with initial investors and the founding students.
Some of the founding students worked for just $2,500 a month,
which is well below the starting salaries for advanced degrees
in electrical engineering. "I get compensated now,"
Broadwater said, "but I have spent 18 months working at
no cost - that's the reality of a start-up where you don't have
much money."
"We almost went out of business," Broadwater recalled.
"We reached a point where we were two months away from running
out of money when we got our first real customer besides EPRI
research.
"It's been very tough timewise," he said. "I spent
most of my waking hours working - a tremendous number of Saturdays
and Sundays. When you have a responsibility like that, you work
all the time. I tried to stay out of that mode early on, but
it fell in my lap. I either had to take charge, or let it fail."
The motivating factor was the technology. "If we hadn't
been involved and able to support it, the technology would have
fallen by the wayside," he said.
In addition to learning a number of important business lessons,
Broadwater said that his involvement in the company has been
a boon to his role as a professor. "For the first time in
my career at Virginia Tech I have a place to meet with my graduate
students. I never really had lab space, and while the workstation
was being developed, I had five graduate students working out
of a single office originally designed for one person. Now we
meet in the company offices; we meet much more often; and my
interaction with the students is better," he explained.
"I've also been exposed to many interesting problems I would
not have encountered otherwise," he said. He cited examples
ranging from utility construction to dealing with hybrid systems
that have been developed over different decades. "For example,
we've had to model some utilities that have a switch that goes
dangerously from phase A to phase C. We have also solved looped
systems where part is three phase and part single phase,"
he said.
"We've run into all kinds of issues with impedance. Utilities
often have as many as 10 three-phase circuits in parallel. Unfortunately,
nobody models it accurately." His research team has obtained
funding to work on that particular problem. "I would never
have been exposed to these issues if I hadn't been involved in
the business."
Perhaps the greatest advantage has been the opportunities made
available to his graduate students. "We have been able to
provide the opportunities of a large academic research group
that would otherwise not have been available with me," he
said. "For example, our work at EDD has exposed us to many
different design calculations that are used and needed by utilities.
These design calculations are applied to the same system and
thus interact with one another. For instance, if you work design
"a" first and then design "b", the results
for the overall design will be different than if you had worked
design "b" first and then design "a." What
is needed is a way for calculations written by different people
to work together for the best solution. Our students have developed
that capability, and have been able to solve problems that they
would never have had the opportunity to solve without the EDD
exposure and support," he explained.
"Would I do it again? Given the advantages to the students
and the industry, yes, I would do it again. I would do it differently,
though: I would do it better."
Not bad for a reluctant businessperson.
|