Inspecting AOVs in Spain8 September 2014
In 2010 the Spanish nuclear regulator launched a new programme, which required all Category 1 safety-related air-operated valves to be tested within three outages. Tecnatom explains how the diagnostics can be carried out.
Across-vendor group of utilities in the USA, the Joint Owners' Group, set up a periodic verification programme for motor-operated valve diagnostics in 2006-7 (see documents ML063470526 and ML071730468 on www.nrc.gov) in response to concerns from the nuclear regulator that static-tested safety-related valves might not function in all design-basis circumstances.
Unlike with motor-operated valve diagnostics, there is not a common regulatory requirement or practice in the field of air operated valve diagnostics in the nuclear sector. First approaches to implement an AOV programme similar to the MOV programme started in 1999, when the Joint Owners' Group released a document (ML020360091) describing the AOV programme proposed by US nuclear power plants to assure the operability of the valves under design basis conditions. It was accepted in March 2000 with comments in RIS 00-003: Resolution of Generic Safety Issue 158: Performance of Safety-Related Power-Operated Valves Under Design Basis Conditions (www. tinyurl.com/jogaov).
In Spain, nuclear facilities must comply with both the regulatory requirements established by Spanish Authorities and with any requirement applicable in the country of origin of the reactor technology. As most of the facilities are either of Westinghouse or General Electric design, the Spanish industry generally has adopted JOG's practices. Thus, the Spanish Electrical Industry Association UNESA which unites Spanish electric power utilities, started the implementation of an AOV programme, based on the JOG publication.
In 2007, the Spanish Nuclear regulator CSN (Consejo de Seguridad Nuclear) published its Technical Instruction on AOV Revision Programmes, mandatory for all Spanish nuclear power plants. This Instruction endorsed the programme presented by UNESA and established deadlines and/or implementation periods for different phases of the programme.
The first phase, to be implemented by December 2009, required several initial activities:
- Scope definition and categorisation
- Set point adjustment
- Revision at component and system level of valves affected
- Issue of any specific "position paper" as needed
- Performance of any activity which may seem, at this stage, necessary to assure valve operability
- Establishment of a testing program which should include diagnostic techniques wherever possible.
The second phase listed the test programme requirements: diagnostic tests on plants AOVs were to start on the first plant refuelling outage from 2010, and full scope of Category 1 valves was to be tested within three outages.
The categorization of the valves was performed individually in each plant by a board of experts. This board included technicians from mechanical and I&C maintenance, engineering, plant operation, as well as external experts, according to the valve programme attributes indicated in ML020360091 and NRC RIS-00-003. The fact that different teams participated and different criteria were applied resulted in very different scopes between facilities of similar design. As an example, two Spanish NPPs of the same design and generation were considered. The scope of Category 1 air operated valves subject to diagnostic test in one of the facilities is of about 12 valves (RHR exchanger flow regulation and by-pass valves, steam dump, chemical and volume control system, main steam relief valves), while the scope in the second one is of about 48 AOVs. In this case, valves of additional systems are considered: auxiliary feedwater, containment isolation, and so on.
This mandatory diagnostic scope of safety-related valves is frequently enlarged with other valves considered critical or important from an operation, plant production or maintenance point of view, valves with high performance requirements (for example, feedwater control valves) or with unexpected malfunction during plant production cycle.
The AOV diagnostic tests are normally programmed and performed during plant outage; 15 to 25 valves are typically tested. In case of malfunction or failure of safety-related Category 1 valves during operation cycle, online testing can also be performed.
During a typical outage, an as-found test is performed, prior to any maintenance and/ or adjustment or calibration of the valve. This as-found signature is analyzed and performance of the valve is evaluated. In the cases where the valve does not comply with engineering requirements, or does not present an acceptable performance, maintenance, calibration and/or adjustment is recommended by the testing technician and analyst. If the valve is subject to any maintenance intervention or adjustment, an as-left test must then be performed in order to assure proper performance of the valve during the next operation cycle.
To perform the testing of the standard scope of 20 valves during an outage, a team of five people is mobilized. Two operators prepare the equipment and install the sensors on the valves; two analysts perform the tests and evaluate the data obtained. A supervisor reviews conclusions of the analysis and exchanges information with plant personnel Information sharing is maintained between technicians and plant personnel for the year to improve valve programme and evaluation.
Testing personnel participating in diagnostic services is qualified under a Tecnatom qualification procedure based on the ASME qualification requirements for inspection personnel. It consists of theory and laboratory practical training, accompanied by field experience.
Since the implementation of AOV Diagnostic Programmes in 2010, Tecnatom has carried out AOV diagnostic services in five Westinghouse pressurised water reactor and two GE BWR units in Spain, as well as in Laguna Verde NPP (Mexico) and in EDF's Blayais, Civaux and Dampierre nuclear plants.
AOV diagnostics are based on measuring valve parameters (pneumatic signals, displacement, electrical and control signals, forces on the stem), acquiring this data through different tests and generating valve signatures to determine performance characteristics parameters.
For AOV diagnostic tests, Tecnatom uses a commercial Teledyne (TTS) data acquisition and diagnostic analysis system, combined with a Tecnatom proprietary development pneumatic interface. This equipment is adapted to test the wide range of AOVs found in nuclear power plants, including linear and rotary, on-off and regulation valves. For the particular case of butterfly valves, that is, no-hard-seated or soft-seated valves, Tecnatom has developed various devices to measure actuator capacity and determine operational margins.
Typically, pressures in different locations of the actuator, stem travel/stroke and loads and control signal are measured using several transducers while the valve is being stroked locally. These data are continuously recorded during different opening and closing strokes, controlled from the data acquisition system, to obtain a valve signature. This allows evaluation of valve performance and validation of valve's capacity and adjustments. Two typical pneumatic diagrams are shown in Figures 1 and 2 for a single-acting diaphragm actuator and for a double-acting piston actuator, with the required signals (in blue). Depending on the configuration of the valve and/or actuator and its associated instruments, fewer signals may be needed.
The measurement interface is typically composed of six pressure transducers with different ranges, to measure all pressures shown in Figs. 1 and 2. These transducers have pneumatic fast connectors to be connected to different pressure points and electrical connections to send the signal read to the data acquisition system. Additionally, current/pressure (I/P) transducers, controlled from the data acquisition system, may be used to stroke the valve and conduct the test without using the valve's control device. This is needed, for example, in the case of on-off valves or pressure-controlled valves, or in the case where a plant I/P converter is suspected of malfunction and the other parameters of the valve need to be independently checked.
Stem travel is measured with a string-pot sensor. It has a voltage output that is recorded in the data acquisition system. Thrust (or torque for rotating valves) is measured on the stem with strain-gauge sensors. Limit switches are monitored if needed.
In order to obtain a more realistic valve signature, measurement devices must not interfere with valve operation. Thus, equipment installation shall be as minimally intrusive as possible. Using calibration points or pressure gauge connections of
the positioner is preferred. Once the test is finished, any modification of plant elements has to be removed.
Up to 16 signals can be acquired and recorded at the same time with the data acquisition system so that all required parameters can be monitored. The system allows any test configuration: ramp test, step test, multi-ramp test, resolution test, frequency-response test, or a combination of these. The most common tests are:
- Decay test. This is the first test done to check the absence of leaks in the actuator.
- Step test. This is used to learn the travel time of the stroke, to check if there is a flow restriction. For example, if a pressure decay is upstream of the positioner, something is restricting the air flow (tubing, regulator, etc).
- Slow ramp test. This is done with a ramp signal travelling from fully closed to fully open, and back to fully closed. (Open-close- open can be done too).
Once the valve signature is obtained, the current condition and performance of the valve can be evaluated.
There are quantitative and qualitative aspects of analysis.
Quantitative analysis consists of comparing the obtained values to the required ones. The plant's engineering services set the critical parameters as well as the acceptance criteria. Although critical parameters may differ between facilities, they typically are: stroke length, seat load and friction load on the stem. Other actuator parameters may be critical too, such as bench-set or spring characteristics for single-acting diaphragm actuators, since the main objective of the diagnostic test is to verify the capacity of the valve to realize its safety function in the most unfavourable conditions. Attention must be paid too to instrument set points and adjustment/calibration tolerances included in plant's valve procedure.
On the other hand, a qualitative analysis of the sensor traces is also performed to consider global performance of the valve. Abnormal signature shapes, non-linearity and atypical curves may be signs of degradation and potential malfunction. Adequate attention must be paid to the qualitative analysis.
An important source of information about valve condition comes from trending analysis. When testing a valve, it is essential to compare new results with previous ones to properly understand valves' operation and, more importantly, to promptly detect any degradation condition.
The diagnostic equipment described is also able to evaluate every kind of I&C device installed on the valve, such as the positioner, regulator, booster, position transmitter, solenoid valves, I/P converters, limit switches, and so on.
AOV diagnostics evaluates a valve's current operating condition without disassembly. Diagnostics allows implementation of condition-based maintenance programmes avoiding typically more extensive preventive maintenance procedures. It provides plant improvements: process efficiency and productivity gains, minimization of valve failures, reduction of spare parts inventory and reduction of labour costs. Furthermore, valve signatures allow verification of valve design criteria, mainly actuators' capability to operate the valve. Finally, trending analysis of signatures detects and characterizes potential degradation mechanisms. ¦
The Air-Operated Valve Users' Group is dedicated to the development and exchange of technical information concerning the performance, design, testing, and maintenance of air-operated valves, compressed air systems, and fluid leak management programmes. Its next conference takes place 14-16 January 2015 at the Hyatt Regency Riverfront Hotel, Jacksonville, Florida. It is intended to improve efficiency, equipment reliability, and safety in industries such as power generation, chemical, oil & gas, naval, and other industrial process control applications. More information is available on www.aovusersgroup.com
Remedios Cervilla, Fernando Vallana and José María Laporta. Inspection and Testing Services Direction, Tecnatom S.A., Av. Montes de Oca 1, 28703 San Sebastián de los Reyes, Madrid, Spain. This paper was based on 'Air-operated valve diagnostics: requirements, programmes and experiences' (A0115), from European Nuclear Congress 2014, 11-14 May 2014, Marseille, France
Case study: Discharge valve of residual heat removal heat exchanger
The actuator is a linear, double-acting piston type, connected to a fish-tail type butterfly valve. The function of this valve is, jointly with the bypass, to regulate the flow through the RHR heat exchanger. Thus it is designed to never be fully closed: the stroke of the actuator is mechanically limited to prevent the valve getting to the fully-closed position, and there is no seat. In the as-found test, in addition to the spaces/clearances typically found in this kind of linear-to-rotary transmission mechanism, the valve presented an important and abnormal increase in the torque signal at the end of the closing stroke and at the beginning of the opening stroke (see Fig. 3).
This abnormally-high-friction torque could also be detected in the enlargement of the mechanical properties plot on the open side of the curve (see Fig. 4). Since the valve has no seat, this friction increase couldn't originate from the friction of the butterfly against a rubber seat. An abnormal point of friction was expected at the origin of the increase of torque.
Major maintenance was recommended, and the mechanical intervention revealed that the butterfly was misaligned in the pipe because of the loss of a set screw, making contact with the walls of the pipe at certain points of stroke. Continued operation in this degraded condition could have led, in the long term, to full blocking of the valve. In this case, the misalignment of the butterfly could only be detected through diagnostic testing. Late detection of this degraded condition could have led, in the worst case, to an unexpected inoperability of a RHR safety valve during plant operation.