AREVA NP White Paper on PRC-24

AREVA acknowledges the need for transient ride through criteria for voltage and frequency and applauds the NERC Standard Drafting Team for tackling such difficult subject matter. AREVA has worked to fulfill Grid Code and Transmission System Operators requirements in other countries where our nuclear power plants are being operated and constructed. These include new EPR plants under construction in Finland and France as well as operating plants in Germany, France and around the world. We also have personnel with extensive experience in generation related NERC Standards activities.We would like to offer to leverage the knowledge gained in these endeavors to support the NERC standard effort.

Draft NERC Standard PRC-24 proposes the standard ride through criteria for voltage shown in Figure 1, which shows the proposed standard for the high and low voltage excursions.

AREVA understands this criteria was developed partly based on a study of the WECC region (reference WECC paper) and the resulting criteria was based on bounding system response for all operating areas, including those where wind generating power plant characteristics are understood to delay voltage recovery once the fault is clearedHowever, this criterion is not technically feasible for large thermal/nuclear generating plants for the following reasons.

  1. NERC standards appear to see generators as a discreet part of the overall bulk power system. While this is accurate for small wind powered induction generators, each large generation plant consists of a generator and a plant auxiliary distribution power system which must be designed and coordinated with the transmission power system.
  2. In the face of an extended voltage decrease and a very slow voltage recovery, as being proposed, the stability of large inertia generators is questionable and has to be investigated.
  3. The assumption of a near to the plant short-circuit failure with a LVRT is not technically correct. The behavior of the auxiliary systems and especially of large MV motors inherent in a steam plant is different when there is a short-circuit and when there is only a voltage drop at the grid connection.

In Order No. 672, FERC has identified 15 points of justification that will be used to analyze standards that NERC proposes for approval. These criteria ensure the proposed standard is just, reasonable, not unduly discriminatory or preferential, and in the public interest. Two of these include

  • Proposed standards must be designed to achieve a specified reliability goal
  • Proposed standards must contain a technically sound method to achieve the goal

AREVA understands the specified reliability goal is identified in the purpose of the standard, which for PRC-24 is:

“Ensure that generators remain connected to the Bulk Electric System during voltage and frequency excursions within the defined criteria of this standard in coordination with other system protection schemes to support transmission system transient stability.”

AREVA believes the current criteria for Low Voltage Ride Through as written misses the target for these two points and thus likely will not be approved by FERC. AREVA also believes the discussion of record on the PRC-24 standard should clearly state that the superposition of extended off nominal voltage and frequency is not technically possible with US equipment design standards and thus not required.

Below, AREVA offers a technical discussion on our concerns with the existing standard approach and offers analternative approach for the SDTs consideration.

The Voltage Ride through criteria curves consists of several diverse parts, as described in the White Paper developed by the Wind Generation Task Force (WGTF) see Figure 2.

Figure 2.
Voltage Ride Trough Boundaries

  1. Normal and Emergency Voltage Conditions – Voltage Tolerance Boundaries

The normal voltage boundaries have been specified to be for the steady-state operating conditions based on the ANSI C84.1-2006 “American National Standard for Electric Power Systems and Equipment – Voltage Ratings (60Hz)”as follows:

  1. Normal Conditions:±5% Continuous Duration
  2. Emergency Conditions:±10%not specified Duration

These Criteria are currently widely used in practice and can be complied with by all types of new generating plants designed with an in-plant voltage regulation capability. In connection with these criteria, all new equipment, both on the transmission system and in new generation plants must be chosen in order to be able to operate and withstand these voltage excursions.

However, many existing power plants were built prior to standards for normal transmission system voltage operating bands and thus were designed without in-plant voltage regulation capabilities. In some cases, assuring nuclear plant offsite power voltage adequacy requires minimum voltage limits very near nominal system voltage. For these plants, close coordination between the transmission system operations and the plant are required to

  1. Select the main step up transformer and auxiliary transformer taps settings to optimize the MVAR support available from the plant under grid conditions expected most of the time.(e.g. When operating on Voltage Schedule as Required by NERC Standard VAR-2).
  2. Assure the transmission system is operated to assure thatsystem voltage near operating nuclear power plants, with the single contingency (plant accident, trip and application of safety system loads) maximum voltage drop, will be adequate to support the safety system operation as required by NRC regulations and new NERC Standard NUC-1.

Thus, requirements for generation to be able to support a wide band of operating transmission system voltage is, in essence, a requirement to back fit these old plants with voltage regulation capability. While this might be the best choice in selected situations, it might be a less cost effective solution in some regions of the interconnection where there are higher concentrations of generation. This also may run counter to regulations in some states, where the utilities make these system design decisions during the long term planning processes and are required by the states to design the system on a least overall fleet cost basis. It is important that the transmission operator and the connected generation fleet cooperate on these issues as now mandated for nuclear plants through NUC-1.

In the course of sighting plants in the US, AREVA has had discussions with some US transmission operators who interpret these draft standards to require the superposition of an extended ability to withstand Emergency Voltage Conditions together with Frequency at 95%. Imposing these conditions on plant auxiliary systems are contrary to motor manufacturing standards in NEMA MG-1 which state

“Section III MG 1-2006LARGE MACHINES—INDUCTION MACHINES Part 20, Page 11

20.14 VARIATIONS FROM RATED VOLTAGE AND RATED FREQUENCY

20.14.1 Running

Induction machines shall operate successfully under running conditions at rated load with a variation in the voltage or the frequency up to the following:

a. Plus or minus 10 percent of rated voltage, with rated frequency

b. Plus or minus 5 percent of rated frequency, with rated voltage

c. A combined variation in voltage and frequency of 10 percent (sum of absolute values) of the rated values, provided the frequency variation does not exceed plus or minus 5 percent of rated frequency.

Performance within these voltage and frequency variations will not necessarily be in accordance with the standards established for operation at rated voltage and frequency. “

AREVA appreciates the basis for Voltage Ride Through Clarification #6 and encourages the SDT to keep this clarification in the final standard.

AREVA designs our new plant power systems with voltage regulation capability to support long-term variations in transmission system voltage, but believes that capitol decisions on when to add plant voltage controls on existing plants should be made on an as needed basis to facilitate improved MVAR support and not to support an arbitrary criterion. A requirement to mandate Automatic Tap changers at existing units when transformers are replaced may be a reasonable long term approach.

  1. LVRT – Three-Phase Fault Clearing Boundary

This part of the curve addresses the “bolted” three-phase fault boundary with normal clearing, as a worst-case scenario for short-circuit fault, after which the generators are required to remain in service.

The short-circuit fault in the proximity of the power plant will be seen from the point of view and response of the power plant similar as the short-circuit fault at the grid connection (switchyard) of the plant. During the time the fault is connected, the generator terminal voltage lowers, depending on the type and distance of the fault to the plant and the generator is able to deliver only a part of the pre-fault active power (MW) to the grid. Since the Turbine torque is still constant, the generator experiences an acceleration proportional tooutput power imbalance. If not cleared within certain time boundaries, which are individual to every plant,it will lead to the loss of the transient stability.

Consistent with FERC expectations, the proposed standard clearing time is 9cycles or 150ms. This is the clearing time of the Zone 1 protection in the connected grid and consists of:

-Relay operation time

-Communication time

-Breaker time

The voltage drop for the duration of the fault is proposed to be down to 0% which is very conservative but acceptable and is consistent with the statement that a near to plant fault will be seen as a fault in the switchyard connecting the plant to the grid.

AREVA believes this is a reasonable basis for the fault clearing time and 150ms should be used for the US standard. The time 150ms is comparable to the clearing time specified in the German Grid Code from E.ON and is smaller than the requirement from NORDEL (Nordic Countries Grid Union) which is requiring the generators to withstand 250ms clearing time of the fault based on their weak grid conditions.

It does not appear that breaker-failure protection times were taken into account as a part of the standards effort. This protection is typically set to “wait” between 0.1 to 0.35 seconds for the main protection to clear the fault and should be based on the critical clearing time for the local generator. This criteria is also correspondent with the definition that the normal design contingencies include among others a permanent three-phase fault on any generator, transmission circuit, transformer or bus section, with normal fault clearing and with due regard to reclosing facilities [Kundur, 1993]. It would seem that a complete standard for transients might address the requirement for the transmission system/switchyard equipment owners to set these breaker failure units and updates the settings as necessary based on the latest stability study results.

Because the required grid protection clearing times needed to assure stability vary, it is again important for the Generation Owners and the Transmission Owners and Planners to collaborate on an on going basis to assure the system design is maintained as needed to assure reliable operation of the bulk power system as additional generation is added as is currently projected.

  1. LVRT – Voltage Recovery Boundary

This is the part of the standard which addresses the system voltage recovery after the fault is cleared. It is important to keep in mind that, while part 2 of the curve can be seen as quasi-static and be stated uniformly for all plants, the expected voltage recovery will vary location by location depending on nearby plant characteristics and its real shape can be determined only by dynamic simulation. For the proposed criteria, the initial condition is that the voltage is still at 0% and linearly recovers to 90% 1.6 seconds after the fault clearing. While this criteria is certainly bounding for all types of generation, it is unrealistic for large generation plants for the reasons stated above.

There is also an important difference between synchronous (mostly used in thermal, nuclear and hydro generating power plants) and asynchronous (wind generating plants)generators that requires the stator of the asynchronous generator to be magnetized from the grid before it works and therefore needs reactive power in order to build up the magnetic field.

The asynchronous generator is able to operate also in a stand alone system, if it is provided with capacitors which supply the necessary magnetization current. It is also required that there is some remnant magnetization in the rotor iron, i.e. some leftover magnetism during the start of the turbine. Otherwise there is a need for a battery and power electronics, or a small diesel generator to start the system). This means that after voltage drops in the supplying grid the asynchronous generator needs relative longer time for supporting the voltage recovery, because it needs the extra time to rebuild its electromagnetic field compared to the synchronous generator.

Synchronous generators are equipped with AVR systems which provide a fault forcing function to help the transmission system recover from faults. With these generators nearby, the system voltage will recover quickly to >90% once the fault is cleared at 150msec.

In this part after the fault was cleared, the leading magnet wheel of the synchronous generator will be slowed down by the recurrent electrical torque and reaches in the stable state under declining oscillations the synchronous rpm based on the grid frequency (60Hz). The proposed standard attempts to describe thisdynamic behavior trough the linearly increasing part of the curve.

If the synchronizing capability of the grid not sufficient to slow down the accelerated magnet wheel after the fault was cleared, the turbo-generator set will transit into slip mode operation, which is associated with enormous mechanical stress. This is in fact the loss of the transient stability.Areva believes the standard should consider this and use a criteria based on ensuring the generator will not reach this undesired operating condition.

Other parameters influence the after fault voltage recovery at the grid connection:

  • The inertia of the turbo-generating set (higher inertia means higher stability)
  • The position of the generator voltage regulator and the resulting operation mode either leading, lagging or cos=1 (measure for the level of stability of the generator prior to the fault=> leading =>max angular displacement =>worst-case)
  • The grid connection voltage level and the related short-circuit current capacity (it is the measure for the resynchronization capability of the grid)

The purpose for this introduction is to show how complicated correlations are in an attempt to represent the entire transmission system by a simple curve of the standard.

According to the WGTF white paper, the voltage points during voltage recovery for Zone 2 three-phase fault with normal clearing time have been drawn into the diagram and a curve was chosen to begin at 0% voltage and 150ms to reach the 90%voltage in a manner that ensures all the voltage points will be within the area specified/marked-off by the curve (Figure 3).

Note that only handful (12) of the voltage points were available for this study which makes it less reliable in terms of statistical evaluation and is unlikely that all possible scenarios are covered. This criteria forces large plants connected to grids with high voltage and higher short-circuit capability (in the Fig.3 represented by the points Nine Mile 230kV (PAC) and Foote Creek 230kV (PAC)) to assure they can ride trough very poor voltage conditions that can not realistically occur on their grid connections.

Figure 3
Existing and new LVTR standard for WECC with Zone 2 Relay Results added

Fig.4 shows a comparison of the proposed standard LVRT vs. a generator stability study from a 1300MW unit which clearly shows how far the proposed boundary and the voltage recovery for the generator terminals after 150ms short-circuit fault differ. In the study the voltage recovery up to 100% is achieved after 900ms versus the by standard proposed 1.6s.

Figure 4
Proposed LVTR standard for WECC vs. 1300MW unit stability study.

The proposed criteria may be acceptable for the generators if they can maintain stability, but likely will cause problems for the auxiliary systems of the thermal and nuclear plants, because of the impact on the voltage in the auxiliary systems. The voltage drop will be propagated through the whole system and the relative slow recovery can cause under-voltage trips for vital consumers which will lead to a plant trip even if the generator remains stable. Even newer plants equipped with On Line Tap Changers can not work against the voltage drop, because the whole event is too fast for the OLTC to react.

AREVA has extensively studied the ability of in-plant power systems to ride through grid initiated voltage transients resulting from the near to plant short-circuit failures. Several slides based on these studies are included in the Fig 5 and 6.

Figure 5

Figure 5 shows the voltage recovery profiles at the MV system distributions within the auxiliary systems of a plant after the clearing of the near-to-plant short-circuit failures, cleared after 150ms and 210ms. These curves are very similar to the voltage recovery profile on the main generator terminals and clearly show the impact of the short-circuit failure to the whole auxiliary system.

This impact is caused by the ratio of the impedances of the equipment which are decisively involved in the Short-Circuit Impedance such as the Step-Up (ZStT), Auxiliary Transformer (ZAUX) and the MV-Motors (ZMot), and determine the resulting short-circuit current in the system.