The Control Room of the Not-Too-Distant Future

This article is the first of a three-part series on the future of hydroelectric control rooms. The two authors blend existing systems with a vision of the exciting possibilities ahead to show how advanced hardware and software will give plant operators seamless control.

By Jery R. Stedinger and Charles Howard

Editor's Note: HRW asked two of hydro's creative thinkers to consider how the control room of the not-so-distant future might look. Next time, how smart technology may soon be making many more decisions on its own.

The sky was gray when Joe arrived for work at Diversified Electric Utility Company's (DEUC) hydroelectric control center. He and Sara were day-shift operators for the utility's six major reservoirs and 12 hydroelectric plants spread across two watersheds. Their assignment was to maximize the value of generated hydroelectric energy, minimize thermal plant costs, protect the environment, and maintain flood protection for several communities.

Joe completed the "hand-over" from the night operator, Jack, and queried the hydro management software, called JENNIFER, to summarize the system status. A colorful block diagram appeared on the 3-foot-by-5-foot screen at the operator workstation, showing levels of flows and reservoir storage.

In her familiar soft voice, JENNIFER began to review conditions of special interest, pausing briefly after each report to see if she should provide more detail or continue on the verbal tour. Joe listened carefully to JENNIFER's concise commentary, sometimes making a verbal request for more information. The large combination monitor and video screen in front of Joe was filled with a changing display of color charts, graphs, and numerical summaries of activities at the various hydro plants. When the tour was completed, the good-natured JENNIFER even asked how Joe's most recent fishing trip had gone. "Great," he responded, delighted to be asked about his favorite recreational activity.

OTTO tracks turbine performance

When the system rundown was finished, Joe asked JENNIFER to yield the active channel to OTTO, DEUC's turbine generating and monitoring software. OTTO recorded and analyzed information on all of the turbines on DEUC's hydro system, including internal temperature distribution, vibration, magnetic field levels, efficiency, and mechanical strain.

Listening to OTTO's unit status summary, Joe saw that a trashrack on the Little Bend Unit 4 was showing larger-than-normal head losses. OTTO already had scheduled downtime for Little Bend Unit 3, based on vibration and magnetic field data showing the unit needed maintenance. Because the Unit 4 problem took precedence, OTTO postponed the Unit 3 work. The software scheduled a maintenance team to clear the Unit 4 trashrack first. OTTO developed optimal dispatches for other units to reduce flow through Unit 4 while meeting the power generation targets set by JENNIFER. The controllers would keep a careful watch on Unit 3 until the postponed maintenance job was performed.

Monitors from DEUC's other hydro plants showed that winding temperature profiles, efficiencies, vibration, and noise levels all were in normal ranges. OTTO tracked these indicators with great care, like a doctor monitoring a patient.

GAIL keeps fingers on weather, flow data

Joe picked up a cup of coffee from the vending machine and went to the control room's second primary console. Sara sat in front of the workstation screen, communicating with GAIL, the utility's metrologic and hydrologic management program.

DEUC's controllers had access to real-time data from a network of rain gauges and from radar and satellite scans across the utility's entire operating region. GAIL employed data from the new SHOOTEM weather satellite, which used both passive and active sensors to scan the hydro system's watershed on a 1-meter-square grid. The link provided detailed readings on snow cover, water equivalencies, temperatures, ground cover, soil moisture, and surface temperature.

GAIL generally used readings from the rain gauges and observed flow rates in small streams to calibrate weather radar returns showing the distribution of rain in the atmosphere during a storm. This morning, GAIL was digesting forecasts provided to the control center over the utility's link to Weather Data Transmission Network Inc., a private vendor who combined and repackaged information from the National Weather Service's NEXRAD (next generation radar) system and private sources. The forecast suggested heavy thunderstorms over much of the region.

GAIL also was concerned with a problem on the streamflow data system. Readings from a gauge on Big Creek did not track correctly, based on flow levels reported from other sites and GAIL's estimate of what flows should be at the gauge. Sara agreed with GAIL's conclusion that system forecasting should ignore readings from the problem site. She also accepted GAIL's recommendation that a maintenance crew depart from its routine schedule to visit the gauge site as soon as possible. Sara made a note to call the crew supervisor with the request as soon as the initial run-through with GAIL was finished — the supervisor liked to get such a special request in person rather than have it show up on his DEUC network terminal.

It was still early in the morning, and low hydroelectric releases during the night had resulted in low streamflows in some parts of the basin. Dissolved oxygen sensors and output from GAIL's water-quality model indicated that conditions were favorable for aquatic life. Biological sensors monitoring the vital functions of several species of fish in special monitoring cages confirmed the calculations. GAIL observed that special music and colored lights used to divert fish away from turbine intakes and toward the available fish ladders and elevators were working perfectly. The biggest threat seemed to be that the big fish would eat the little ones.

Later in the day, when higher temperatures and algal oxygen demand began to reduce oxygen levels, streamflows would be increasing. If necessary, some water releases could be diverted through re-aeration weirs or baffles in the larger reservoirs that would direct surface water into the turbines to improve oxygen levels. Priority would be given to reaches viewed as safe havens for fish and spawning. GAIL would monitor the situation and sound the alarm if environmental modeling suggested it was necessary, or if monitored species showed signs of stress.

On the water, a stubborn angler

Joe heard JENNIFER summoning him to the workstation. When releases were being stepped up at the Shale Creek plant, JENNIFER's sensors detected what appeared to be a person remaining in the channel below the plant. Returning to the workstation, Joe brought up a video image of the area. JENNIFER zoomed the video camera in on a fisherman. Uncertain whether the lingerer had not heard the alarms or was just too stubborn to heed them, Joe punched a button that patched his microphone to the reservoir's supplementary public address system. There was no response from the image on the screen. He made a note to have the maintenance crew check the speakers and directed JENNIFER to increase flows slightly at the site to send the fisherman a message.

JENNIFER instructed other plants to provide the scheduled power; OTTO shuffled generation schedules. When the higher flows at Shale Creek prompted the fisherman to join others on shore, JENNIFER and OTTO readjusted the system again.

With operations back to normal, Joe returned to writing his portion of an operating analysis for a state utility commission filing on DEUC's Dos Ojos plant. The manager of the utility's hydro division had asked Joe to write the technical analysis for the report. It was easy: JENNIFER generated the data charts and tables for all the required scenarios and Joe had to supply only the narrative and commentary. JENNIFER checked Joe's analysis and edited the text. The report then went to the corporate offices, where the legal department wrapped it in a pretty package for presentation to the state utility commission.

By late morning, early-afternoon thunderstorms appeared more and more likely. JENNIFER worked through several options for using the extra power. Over a link with the database at DEUC's main dispatch unit, the hydro control room collected information on the status of the utility's thermal units, pumped-storage facilities, and spot market energy prices. The analysis also considered utility controllers' latest projections of power values for that day and updated estimates on production cost and power availability for neighboring utility systems and independent power producers.

Power prices had not risen to their afternoon high, and the upper reservoir at the utility's Big Tree pumped-storage project still was full from the overnight pumping schedule. Thus, the excess power produced by JENNIFER's adjustments would be sent to DEUC's hydrolysis units to generate hydrogen and oxygen for the energy market or later combustion.

Rapid runoff raises risk of spilling

Sara asked GAIL to make a detailed analysis of the risk of spilling at critical plants, using flow models and up-to-date satellite weather and soil data. GAIL's analysis indicated that, because the basin already was wet, significant rainfall would run off rapidly and overwhelm forebays at several plants. Adjustments already had been made in flood-control storage at some of the system reservoirs.

As Sara scanned the spill analysis on the screen at her workstation, Joe wandered over and launched into his standard lecture on how fortunate she was to be working with GAIL. In the old days, he said, DEUC had been lucky to have a reliable forecast for 12 hours ahead, much less three or four days, or a week.

Joe's "old days" were before integrated climate/hydrology systems like GAIL, of course, and some mathematician's discovery of an inverse convolution catastrophe theory that allowed development of good eight-day forecasts of the arrival and intensity of major frontal storms. Although SHOOTEM generated better information on hour-to-hour developments, the National Weather Service and private forecasters still had trouble forecasting the behavior of summer convective thunderstorms.

With a last shot at Sara about the skill it took to manage hydro in the days before GAIL, Joe returned to his station. JENNIFER had called him to review the revised schedule for the day. JENNIFER could have alerted him sooner but did not want to worry him with the details until she and OTTO had worked them out. Joe could take charge now.

Joe was surprised to note JENNIFER was not using the full capacity of the Hawk Canyon plant. JENNIFER reminded him that releases at the plant were being constrained to protect downstream construction. "These computers," Joe laughed to himself. "They always take themselves so seriously, but they don't know everything." He had driven past the downstream construction site on the way to work that morning and noticed the crew loading machinery on flatbed trucks. Joe called the construction superintendent to confirm his observations — and mildly chide the man for not reporting the changed status earlier. He then directed JENNIFER to revise the schedule.

By early afternoon, Joe had finished his portion of the utility commission filing, and JENNIFER was reviewing the document. Joe turned his attention to the annual maintenance summary, which was due to division management by the end of the week. OTTO produced the statistics and colorful graphs, but Joe had to polish the otherwise-canned text and provide an explanation of unusual events.

The plan unfolds

Joe watched conditions develop just as he and JENNIFER had planned. Units automatically ramped up to peak levels, and forebays were drawn down to their lowest operating points. Then, the forebays began to fill as releases from upstream arrived and natural flows began to increase from the rain. GAIL continued to update her forecasts, monitoring readings on streamflow levels and weather data arriving over the weather data link. Water ran off the watershed and filled creeks and streams. Eventually, stream channels filled throughout the basin.

Joe felt like a conductor of a well-rehearsed orchestra. He mused over the time when hydro system operators really struggled on days like this, scrambling to make changes manually or selecting a unit to follow load. Those days seemed incredible compared to the current system. Today, he and JENNIFER set the system targets … they picked the tune. Then OTTO played the system like a musical instrument, smoothly moving toward the targets without sudden changes, missed beats, or sour notes. In particular, some units could respond faster to loads, but their generation was gently replaced by that of more economical units.

 Click to Enlarge
 This control room illustrates the degree of computerization that is possible today for hydroelectric facilities. Using advanced hardware and software, plant operators can seamlessly control multiple facilities.

Joe thought back to a training exercise he and Sara had gone through the week before. Both had replayed the meteorological data and inflow sequence of the previous month's storm combination. Rainfall had been very heavy in the northern part of the region that day, resulting in an unnecessary spill at the Palos Verde plant. Operators at the time had drawn down the two southernmost reservoirs more than necessary, resulting in an avoidable loss of head. After several tries, Sara and Joe had showed they could operate the system to generate more electricity and waste less water. Such exercises made Sara, a relatively new operator, feel more comfortable with how the system worked, what to watch for, and how things looked when they were not quite right.

Even the boss can have fun

Joe heard the signal for an incoming call over the DEUC communication network. He directed JENNIFER to accept the call, and the colorful system summary on his station screen dissolved into a 3-foot-by-5-foot image of his boss, Wayne, and a companion. The hydro division chief was at a university conference center in Toronto, Ontario, Canada, for a workshop on remote, multiple-turbine control systems. Wayne explained that he and an administrator from the Geheyan hydro plant in China had been swapping storm-management stories. He asked Joe to open access to JENNIFER and OTTO so the discussion could continue with support from the DEUC database. Joe authorized the access.

Wayne took the opportunity to rattle off a list of items and ideas he wanted Joe to pursue. As Wayne talked, JENNIFER displayed a summary of the requests in the corner of the screen along with her own observations: This one she already had done, another was more difficult, and the last she did not think was a good idea at all. Thank goodness Wayne could not see JENNIFER's comments.

As the end of the shift approached, Joe dictated a couple of entries in the log and reviewed the status list of pending problems and required maintenance. As conductor, Joe knew he should make sure all the instruments in the orchestra were in good condition. He was just finishing the review when Pablo, the lead second-shift operator, arrived. The two went through the hand-off protocol, and Joe said his "goodbye." As he walked to the control room door, Joe heard Pablo switch to Spanish and instruct JENNIFER to run through an initial system summary.

Driving home, Joe was pleased by how well he and his orchestra had performed. He thought he might still get in a little fishing after dinner. GAIL had told him where the fish were biting.

Jery Stedinger, PhD, is a professor at the Cornell University School of Civil and Environmental Engineering. Chuck Howard is a principal in CddHoward Consulting Ltd.

More HRW Current Issue Articles
More HRW Archives Issue Articles

To access this Article, go to: