Reclamation’s Upper Colorado Region Water Office makes monthly water release schedules in acre-feet for Glen Canyon based upon anticipated inflows to the lake, release requirements to the lower basin and storage targets for the lake. This monthly water schedule is passed on to WAPA’s Resource Scheduling Office in Montrose, CO. WAPA’s scheduling office then develops an hourly release schedule. This schedule takes into account Glen Canyon ROD requirements and the requirements and agreements made by WAPA and Reclamation regarding the operation of the other CRSP powerplants. WAPA’s dispatch office in Phoenix, Arizona then provides Reclamation’s Glen Canyon Power Operations Center with a daily, 24 hour operating schedule, identifying WAPA’s desired hourly electric power schedule (in megawatts). This schedule is then put into the computer system at Glen Canyon by Reclamation (2). This schedule then becomes the megawatt quantities which are targeted during the “real time” operation at Glen Canyon Dam.

The Glen Canyon Power Operations Center operators utilize the generation schedule supplied by WAPA to operate the generating units while giving consideration to efficiency and equipment characteristics. In addition, the generators automatically respond to a “regulation signal” developed and electronically transmitted to the dam by WAPA for continuous response to power system load and frequency changes. (This is further described later in this paper).

Shortly after Interim Flows began (August , 1991), WAPA met with staff at the Glen Canyon Environmental Studies (GCES) and with other Reclamation representatives. WAPA requested flexibility in order to regulate an electrical control area. This resulted in an agreement whereby WAPA was allowed to fluctuate 10% around any operational constraint and that WAPA would meet with GCES staff monthly to discuss any operational issues that surfaced. Thus, the Interim Flow downramp requirement of 1,500 cfs could be as high as 1,650 cfs and still be in compliance (1,500 cfs + 150 cfs).

After the ROD and Operating Criteria were signed, WAPA and Reclamation negotiated a MOU to more fully detail operational issues. Reclamation was interested in changing the 10% band width because it believed that the ROD and Operating Criteria didn’t specify that kind of flexibility. Instead, WAPA was allowed to use up to 1,000 cfs for regulation purposes. Compliance was to be monitored by SCADA using a one hour integrated value. It was thought that, even though a downramp at any one moment of time could be as high as 2,500 cfs (1,500 cfs + 1,000 cfs), it would be uncommon for the downramp to exceed 1,500 cfs since this is recorded as an integrated value over the hour.

2The EIS identified Emergency Exception Criteria as common elements in all alternatives. They are defined in the EIS as follows (3):

“Normal operations described under any alternative would be altered temporarily to respond to emergencies. NERC has established guidelines for the emergency operations of interconnected power systems. . . . Examples of system emergencies include:

• Insufficient generating capacity
• Transmission system: overload, voltage control, and frequency
• System restoration
• Humanitarian situations (search and rescue)”

When emergency exception criteria are invoked, normal operations are suspended until the emergency has ended, or WAPA has discharged its North American Energy Reliability Council (NERC) responsibility regarding the emergency. Emergency exception criteria are further defined in the operating criteria and in the MOU on Glen Canyon operations signed by WAPA and Reclamation.

It was the invocation of emergency exception criteria on April 1st, 1998 which piqued the interest of some AMWG members in Reclamation and WAPA’s interpretation of the ROD.

As a control area operator, WAPA “regulates” the transmission system within a prescribed geographic area. WAPA is required to react to moment-by-moment changes in electrical demand within this area. Regulation means that “automatic generation control” will be used to adjust the power output of electric generators within a prescribed area in response to changes in the system frequency, time error, and tie-line loading, to maintain the scheduled level of generation in accordance with prescribed NERC criteria. The “record” used to calculate the degree to which WAPA is responding to these change on the transmission system is called the “ACE” – Area Control Error.

Hydro facilities such as Glen Canyon have an inherent design that allows them to respond rapidly to changes in power system demands. Other control area operators, across the nation, that do not have hydropower plants must either build small units (natural gas or oil-fired being the most likely) or add this capability to the design of larger units.

The transmission system that WAPA distributes power through is dynamic. Load requirements are constantly changing as a result of either the demands of the customers connected to it or changes that occur in other interconnected power systems. WAPA maintains the ACE signal to record its response to the fluctuations in system “loading”; (an effort to maintain a balance between power being consumed and power being generated – described above). If more demand is placed on the transmission system than is being generated, the resulting ACE is negative. Generators automatically respond to this condition by increasing generation. If demand is less than generation, ACE is positive. Generators automatically respond to this condition by reducing generation. The targeted ACE is zero.

The ACE signal that is sent to Glen Canyon where it is effectively added to, or subtracted from, the existing scheduled hourly generation base point. Therefore, at any moment during the day or night, Glen Canyon might be producing more or less power than the current hourly megawatt schedule.

The ACE signal is transmitted to Glen Canyon Dam every four seconds. The NERC requirement for regulation is that the ACE must “cross” the zero target every 10 minutes. The frequent “swings” in generation are described in the MOU signed by Reclamation and WAPA (4):

“These changes which occur many times during the hour are both positive and negative in relation to the schedule. The resulting output from Glen Canyon generators is an envelope of generation swings that are frequent, small in magnitude, the average of which approximates the original schedule.”

Both emergency exception criteria and regulation have been described in this paper. Recently, some members of the AMWG and TWG have focused their concerns and discussion on “emergencies”. However, the issue at hand and the subject of this paper is regulation. It is largely regulation that has resulted in the occasional exceedance of the 1,500 cfs downramp.

The target values set in the SCADA are usually not exactly achieved. Moreover, the water released from the dam is not precisely measurable. The reasons for this are explained below.

As per the ROD, the SCADA system is the official means of reporting flow conditions at Glen Canyon. There are currently two methods by which water passing through the power plant at Glen Canyon is measured and reported to SCADA.:

Turbine rating curves: The turbine curves describe the performance of the turbines and were used to develop a data base in the SCADA system at Glen Canyon for reporting operations and water releases. The use of the SCADA system for reporting water flow was the method in place at the time the ROD was implemented. The turbine rating curves were derived from testing the turbines and generators and take into account the number of megawatts being generated, the amount of head, and the unit efficiency. The turbine curves are considered to be accurate to within ±3%. Frequent periodic samples are taken on the amount of generation, and the results are converted and integrated by the SCADA system, into an hourly flow number in (cfs).

Accusonic Flowmeters: Each generator penstock has several transducers attached to them to measure the actual instantaneous flow through that penstock. The flow meters are considered to be accurate to within 1% when reporting directly to the Accusonic recording instrument. The flows from all penstocks are sampled every 15 seconds, accumulated by the flowmeter computer and then averaged to show what the flow has been over the last hour.

This average is also fed to the SCADA system. To condition the signal from the Accusonic flowmeters for use/input to the SCADA system, the accuracy is considered to be about ±3%. Presently, these flowmeters are not being used as the official method of reporting, as efforts are being made to verify the accuracy of converting the flowmeter data.

The relationship between water flow through a unit and the amount of power generated is neither constant nor linear. There are several variables that affect the amount of water flowing through the power plant at Glen Canyon:

WAPA’s Preschedule: In determining generation releases, WAPA preschedules the power plant using a conversion factor to translate energy (megawatts) to water (cfs). A single conversion factor is used for the entire month based on the power/water ratio from the previous month. This method is different from the SCADA report, which reflects the turbine rating curves. By comparison, this measurement differs from the Accusonic flowmeters by approximately ±7%.

Lake elevation: The difference in elevation between the forebay and the tailbay is referred to as the “head”. “Head” is affected by lake elevation and the amount of water being released from the generating units. Changes in “head” affect the efficiency of the generating units and head may differ over the course of the day and from one month to the next. WAPA’s scheduling office uses the “head” from the previous month in preparing its schedule. Because of this, and the possibility of “head” changing over the day, the generation at Glen Canyon will always result in water releases that differ from anticipated or scheduled release.

Generator efficiency: Generator efficiency is relatively constant while turbine efficiency varies significantly with discharge and head.

Starting or stopping: When a unit is shut down, there is water flow through the turbine even though there are no energy is being generated. A unit shutdown, or one in standby status; waiting to be started, has wicket gate leakage of approximately 50 cfs. Also, when a unit is started or stopped, water is released before an electricity is generated. As an example: in preparation to bring a unit on line to meet WAPA’s schedule, water release begins and “spins” the generator for a 10 minute period to allow the unit to come up to normal operating temperature. This will add about 155 cfs to the total water release during that 10 minute interval. The water released during this period is not “counted” in the quantity of water released in the hour. The water that is “counted” is that water recorded by the SCADA, which is the official measurement for compliance with the ROD.

Number of generators: Both WAPA and Reclamation have historically used 60 megawatts as the target for the river downramp limit of 1,500 cfs and 160 megawatts as the target for the river upramp limit of 4,000 cfs. Generation schedules supplied by WAPA often reflect the maximum downramp or upramp limits from hour to hour. If a request for generation at Glen Canyon provides for an upramp of 160 megawatts in the next hour and the control center operators need to bring another unit on line, water release could exceed the upramp river constraint by 155 cfs for a short period of time. 155 cfs is 3.8% of the upramp river limit.

Lee Ferry Gage: Another report of flow conditions out of Glen Canyon Dam is the USGS gage at Lee Ferry, approximately 15 miles downstream. This is the official point of measurement for water deliveries to the lower basin states. Gages also report with error. The USGS gage at Lee Ferry is rated as “excellent”. An excellent rating is given to a gage if it reports within an error range of ±5%. The gaging station reports flow every 15 minutes.

Additional data may be necessary to quantify impacts, if any, of small deviations resulting from regulation. There is, however, some evidence that the environmental impact of regulation does not differ from those described for the preferred alternative in the EIS. In an article published in “Rivers”, an analysis was made of the effects of regulation and other electrical reliability criteria on the river ecology of the Lower Flathead River in Montana. The authors concluded that “ Very small fluctuations around the hourly mean discharge [resulting from operations of the Kerr Dam to meet electrical reliability criteria] have little, if any, quantifiable effect on the downstream aquatic environment because short-term electrical excursions resulting in flow fluctuations of short duration are quickly dissipated by river hydraulic processes. (5)” This conclusion seems to be borne out through analysis of the data regarding flows below Glen Canyon Dam. In June, 1998, while the downramp, as recorded by SCADA, exceeded 1,500 cfs about 4% of the time, there is no exceedence at the Lee Ferry gage. River attenuation between the dam and Lee Ferry reduces fluctuations. In fact the highest down ramp recorded at the Lee Ferry gage during this month was 1,216 cfs. Since the ROD limitations on the downramp were established in order to reduce the degradation of beaches in the Marble Canyon area (below Lee Ferry), one could use the conclusion from the “Rivers” article as a working hypothesis for any further investigations in this area.

Choice of Measuring Device for Compliance with the ROD

As described above, different measuring devices exist to measure the actual water releases from Glen Canyon Dam. They all measure with a lesser or greater degree of error. Under the USBR/WAPA MOU, the official record of compliance is the SCADA. This could be changed by joint agreement of the two agencies.

Once a measuring device is established, a “tolerance level” should also be established due to the inability of the operators of the dam to precisely control flows (this is an issue regardless of the conclusion regarding “regulation”, explained below).

Because of changing conditions such as generator unit efficiency, generator loading and transmission system fluctuations, it is difficult to accurately control releases from Glen Canyon Dam. It is not as simple as just turning off a switch and no flow occurs. If the river limits are to be considered hard limits, Glen Canyon will not be able to participate in electrical system regulation as fully as if the limits were “target” constraints. In order for Glen Canyon to deliver scheduled generation and to fully participate in power system regulation with WAPA, the down ramp limit of 1,500 cfs and the upramp limit of 4,000 cfs must be hourly averages to achieve, not hard inviolable limits.

An alternative is to limit the upramp to 3,000 cfs and the downramp to 500 cfs. This change would allow for 1,000 cfs of operational flexibility. This flexibility would be used for transmission system regulation. This would be the only way to more fully ensure that the ROD- specified flows are not exceeded while allowing Glen Canyon Dam to participate in transmission system regulation. However, this alternative would severely limit the capability of Glen Canyon Dam to follow electrical demand on an hourly basis.

If ramp limitations, as specified by the ROD, are hard constraints, either Glen Canyon Dam does not participate with WAPA in regulating a load control area to the extent it has been, or ramping to follow hourly load is much more severely limited than that described in the ROD. Is it reasonable to interpret the ROD as requiring this choice?

The impact of regulation, as differentiated from the impact of the “preferred alternative”, should be analyzed and/or monitored. This may involve simply paying greater attention to the flows at the Lee Ferry gage and a broad dissemination of these data. On the other hand, it may require an analysis (by USBR, GCMRC or some other party) of specific natural resources in between the dam and Lee Ferry, in the Lee Ferry area , and in the Marble Canyon area. The AMWG would likely specify the degree of analysis desired.

It does not seem reasonable that the ROD was intended to require either: 1) the exclusion of Glen Canyon Dam from participating in regulating a load control area, or 2) participation in regulating a load control area but limiting hourly load following to 500 cfs downramp (or something like it). Further, the environmental impact of regulation appears to be minimal, based on a review of the flows at the Lee Ferry gage. Finally, it has become clear to Reclamation and WAPA that precise and inviolate control of water releases from Glen Canyon Dam is beyond the capabilities of the machinery and electronic controls. For these reasons, Reclamation and WAPA intend to impose the following changes:

1. Limit scheduled down ramps to 1,450 cfs (rather than 1,500),
2. Employ the use of SCADA table in the preparation of WAPA’s hourly schedules,
3. Investigate the transfer of a portion of the “ACE” to other CRSP powerplants,
4. Move the ACE signal back to Glen Canyon Dam (i.e. return “regulation” to the dam) once it has been demonstrated that the ACE will not cause “routine” excedences,
5. Work with the AMWG to determine the proper level of analysis required to address the potential downstream impacts of regulation in Glen and Grand Canyons,
6. Educate all interested parties on the power system and equipment operations and the associated results to water releases at the dam,
7. Add release data to WAPA’s web site and provide periodic reports to the TWG on the operation of Glen Canyon Dam and,
8. Continue efforts to for better coordination between Reclamation and WAPA regarding the scheduling and operation of Glen Canyon Dam.

(1) US Department of Interior, 1995. Operation of Glen Canyon Dam Final Environmental Impact Statement. US Department of Interior, Bureau of Reclamation, Upper Colorado Region, Salt Lake City, Utah. Hereinafter referred to as: “EIS.”

(2) This computer “system” consists of hardware and software and has the name of Supervisory Control And Data Acquisition (SCADA).

(3). Emergency exception criteria are described in detail in Attachment 6 of the EIS.

(4) USBR/WAPA Interagency Agreement – July 7, 1997

(5) Jourdonnais, Jon H, & Hauer, Richard F., Electrical Frequency Control and its Effects on Flow and River Ecology in the Lower Flathead River, Montana, Rivers, Studies in the Science Environmental Policy and Law of Instream Flow, Vol 4, Num. 2, April 1993, S.E.L. & Associate

This web document was adapted from a CRSP MC white paper dated January 1999.