NB/Z 20326-2014Analysis methodology for uncertainty of setpoints for instrumentation of safety system in nuclear power plants (English PDF)

Foreword

This document was issued on 15 October 2014 by the National Energy Administration of the PRC and takes effect on 1 March 2015.

It is a NB/Z guiding technical document: it does not oblige, it guides.

It is classified under ICS 27.120.20, Chinese classification F 65.

This guiding technical document was drafted in accordance with the rules given in GB/T 1.1-2009.

This guiding technical document was prepared with reference to ISA-RP67.04.02-2000 Methodologies for the Determination of Setpoints for Nuclear Safety-Related Instrumentation.

This guiding technical document was proposed by the Nuclear Power Standardization Technical Committee of the Energy Industry.

This guiding technical document is under the administration of the Nuclear Industry Standardization Research Institute.

Main drafting organization of this guiding technical document: Shanghai Nuclear Engineering Research and Design Institute.

Main drafters of this guiding technical document: Chen Yanhui, Gu Qingwen, Jiang Lijun, Lu Shudong, Liu Yong, Li Xiaoyan, Ma Xusheng, Wu Xueqiong, Xu Dongling, Zhu Liang.

Introduction

The main subject of this guiding technical document is the use of generally accepted methods to calculate the setpoint of a single instrument channel. The use of these methods is of great importance in guaranteeing the normal operation of the plant, in enveloping the accident analysis, and in guaranteeing the integrity of the fission product boundary.

The purpose of this guiding technical document is to give the person performing the calculation a definition of the various areas that have to be considered during the calculation of instrument channel uncertainty, and to offer some recommendations on the method of determining the setpoint.

Many concepts drawn from probability theory and mathematical statistics are used in this guiding technical document. Since those subjects are not the object of this document, the reader is advised to read first some material on the terms used here and on the related basic concepts of mathematical statistics, in order to establish a grounding in those areas.

In addition, this guiding technical document recognises that some of the safety-related setpoints defined by a plant are not closely related to the safety analysis, and that from a system point of view there may be no clearly defined limiting value at all. In such cases a graded approach may be used to classify the safety-related setpoints of the plant, the grading being based on the contribution made to plant safety. On the basis of such a classification it is then possible to give guidance on the determination of channel uncertainty. How the specific criteria for the grading are established, and the method of dividing the levels used for the analysis, are not discussed in this guiding technical document.

1 Scope

NB/Z 20326-2014 is the Chinese guiding technical document that sets out how to calculate the uncertainty of a trip setpoint in the safety instrumentation of a nuclear power plant. It exists because a reactor trip setpoint is not a number an engineer chooses: it is the end of an arithmetic chain that begins with the safety limit the plant must never cross. The safety analysis fixes an analytical limit below that safety limit, and the trip has to be set far enough below the analytical limit that the protection still acts in time even when every instrument in the chain is at the unfavourable end of its tolerance at once. Set the trip too close to the analytical limit and the plant can cross the limit while the instrumentation still reads normal; set it too far away and the reactor trips on noise and the plant is unavailable. Everything therefore turns on knowing honestly how wrong the instrument channel can be. This document builds that number step by step: it defines the twenty-nine terms the calculation uses, from as-found and as-left state through bias, drift, span and reference accuracy to channel uncertainty itself; it requires the channel to be drawn as a configuration diagram from the process interface to the bistable, with the environmental conditions each part experiences; it enumerates the sources of uncertainty part by part, separating process measurement effects such as impulse line temperature and fluid density from instrument effects such as temperature, pressure, drift and power supply variation; it classifies each contribution as random, bias or abnormally distributed, which decides whether it may be combined by the square root of the sum of the squares or must be added arithmetically; and it then derives the trip setpoint and the allowable value used during periodic testing. Nine informative annexes carry the worked detail, covering differential pressure level measurement, flow accuracy, pipe head loss, RTD accuracy, insulation resistance, statistical analysis, digital signal conditioning and plant-specific calibration data. The document was prepared with reference to ISA-RP67.04.02-2000, and is the companion to NB/T 20072-2012, which governs the determination and maintenance of the setpoints themselves. It was issued on 15 October 2014 by the National Energy Administration and took effect on 1 March 2015.

This guiding technical document specifies the methodology for analysing the uncertainty of the trip setpoints of safety system instrumentation in nuclear power plants.

This document applies to the analysis of the uncertainty of the trip setpoints of safety system instrumentation in nuclear power plants.

2 Normative references

The following document is indispensable for the application of this document. For dated references, only the edition cited applies to this document. For undated references, the latest edition, including all amendments, applies to this document.

NB/T 20072-2012 Determination and maintenance of trip setpoints for safety system instrumentation in nuclear power plants

3 Terms and definitions

The following terms and definitions apply to this document.

3.1 allowable value

A limiting value that the trip setpoint may take during periodic testing; if this limit is exceeded, appropriate measures are taken. [NB/T 20072-2012, definition 3.1]

3.2 analytical limit (of setpoint)

A limiting value of a measurable or calculable variable, determined by the safety analysis, that ensures the safety limit is not exceeded. The margin between the analytical limit of the setpoint and the safety limit shall take account of: the response time of the instrument channel; and the transient range of the postulated accident. [NB/T 20072-2012, definition 3.2]

3.3 abnormally distributed uncertainty

An uncertainty that does not have the characteristics of a normal distribution.

3.4 as found

The state of an instrument channel, or of part of a channel, after one operating cycle and before recalibration, if recalibration is necessary. [NB/T 20072-2012, definition 3.3]

3.5 as left

The state of an instrument channel, or of part of a channel, after calibration or verification of the setpoint of the terminal setpoint device. [NB/T 20072-2012, definition 3.4]

3.6 bias

A component of uncertainty that always carries the same algebraic sign, positive or negative, and is expressed as a bounded error of a given magnitude.

3.7 bistable

A device or component whose output state changes as soon as the signal reaches the set value.

3.8 dependent uncertainty

Components of uncertainty are dependent on one another if, for known or unknown reasons, a significant correlation exists between them. Such correlation is typically produced by a common cause.

3.9 drift

An undesired change in output occurring over a period of time, where that change is unrelated to the input, the environment or the load. [NB/T 20072-2012, definition 3.5]

3.10 effect

The change in output produced by a change in external temperature, pressure, humidity, radiation or similar conditions.

3.11 error

The algebraic difference between the measured value of the measurand and its true value. [NB/T 20072-2012, definition 3.6]

3.12 independent uncertainty

Components of uncertainty are independent of one another if neither their magnitude nor their sign is significantly correlated. [NB/T 20072-2012, definition 3.9]

3.13 instrument channel

An arrangement of components or modules, including the sensor, that produces a single protective action signal as required when the conditions of the nuclear power plant demand it. [NB/T 20072-2012, definition 3.10]

3.14 margin

An allowance added to the instrument channel uncertainty when the setpoint is determined. The margin increases the difference between the setpoint and the analytical limit.

3.15 module

An assembly of interconnected components that forms a single device, instrument or piece of equipment, which can be disconnected, removed and replaced by a spare as a unit, which has fixed functional characteristics, and which can be tested as a unit. Provided it meets this definition, a module may be a printed circuit board of a large device, a withdrawable circuit breaker or another sub-assembly. [GB/T 13284.1-2008, definition 3.14]

3.16 nuclear safety-related instrumentation

Instrumentation necessary to accomplish the following functions: a) control of reactivity; b) removal of core heat; c) containment of radioactive material and control of operational discharges, and limitation of accidental releases.

3.17 primary element

The system element that quantitatively converts the energy of the measured variable into a form suitable for measurement.

3.18 random

A variable whose value cannot be predicted accurately after a period of time and can only be estimated by means of a probability distribution function. Note: the word random as used in this document is shorthand for approximately normally distributed at random. A randomly and normally distributed uncertainty is equal in its positive and negative extent relative to the median. The random uncertainty components of the process measurement assembly, the signal conditioning assembly and the trip actuation assembly belonging to the same measurement channel may all be combined by the square root of the sum of the squares method. [IEEE 100-R-2000]

3.19 reference accuracy

A defined limiting value which the error does not exceed when the device is used under the specified operating conditions. [NB/T 20072-2012, definition 3.13]

3.20 safety limit

A limiting value specified for an operating parameter within which operation of the nuclear power plant is safe. [NB/T 20072-2012, definition 3.14]

3.21 sensor

The part of the measurement channel that responds to a change in a plant process variable or condition, and that converts the measured process variable into an electrical, optical or pneumatic signal. [IEEE 603-2009]

3.22 signal conditioning

One or more components that perform signal conversion, buffering, isolation or mathematical operations as required.

3.23 signal interface

The physical medium, such as cables and connectors, used to transmit the process signal between the process measurement assembly, the signal conditioning assembly and the terminal trip actuation assembly of the instrument channel.

3.24 span

The difference between the upper and lower limits of the measuring range. Note: modified from GB/T 17212-1998, definition P1.1.0.15.

3.25 test interval

The time elapsed between the start, or the completion, of two tests of the same kind carried out on the same sensor, instrument channel, load group, safety system or other specified system or device. [NB/T 20072-2012, definition 3.16]

3.26 tolerance

The permissible deviation from a specified or true value. [IEEE 100-T-2000]

3.27 trip setpoint

The terminal setpoint specified in advance in order to initiate a protective action. [NB/T 20072-2012, definition 3.17]

3.28 measurement uncertainty; uncertainty of measurement

A non-negative parameter which, on the basis of the information used, characterises the dispersion of the values attributed to the measurand; abbreviated to uncertainty. Note 1: measurement uncertainty includes components arising from systematic effects, such as the components associated with corrections and with the values assigned to measurement standards, and the definitional uncertainty. Sometimes estimated systematic effects are not corrected for but are instead treated as components of uncertainty. Note 2: this parameter may be, for example, a standard deviation called the standard measurement uncertainty, or a stated multiple of it, or the half-width of an interval having a stated coverage probability. Note 3: measurement uncertainty generally comprises several components. Some of them may be evaluated by Type A evaluation of measurement uncertainty from the statistical distribution of a series of measured values and may be characterised by standard deviations. The other components, which may be evaluated by Type B evaluation of measurement uncertainty, can also be characterised by standard deviations obtained from probability density functions based on experience or other information. Note 4: in general, for a given set of information, the measurement uncertainty is associated with a stated value attributed to the measurand; a change in that value results in a corresponding change in the uncertainty. Note 5: this definition is given according to the 2008 edition of the VIM, whereas the definition given in the GUM is: a parameter, associated with the result of a measurement, that characterises the dispersion of the values that could reasonably be attributed to the measurand. [JJF 1001-2011, definition 5.18]

3.29 channel uncertainty

The quantity used to express the uncertainty of the output of an instrument channel caused by uncorrected random errors or systematic errors, or the margin adopted for that purpose. Channel uncertainty is normally expressed as a probability or a confidence level. [NB/T 20072-2012, definition 3.18]

4 Symbols

The following symbols apply to this document.

A/D: analogue to digital

D/A: digital to analogue

R: combined random uncertainty

REQ: equivalent resistance

RT: total resistance

RTD: resistance temperature detector

STD: controlled, standard reference accuracy

TDF: turndown factor, the upper range limit divided by the calibrated span of the device

URL: upper range limit, the maximum measurement the device permits

5 Preparation for determining the instrument channel setpoint

5.1 General

The discussion that follows recommends a series of steps and an order for the calculation of instrument channel uncertainty or setpoint analysis. Its purpose is to provide the basic instrument channel configuration and functions, together with the sources of uncertainty that may arise, so that a detailed discussion of the specific elements can be carried out.

5.2 Configuration diagram of the instrument channel

A configuration diagram of the instrument channel is drawn so that every part of the channel, from the process interface through to the bistable, is identified together with the environmental conditions each part experiences.

5.3 Determination of design parameters and sources of uncertainty

The environmental effects and the assumptions associated with the environmental conditions are given in 6.2.5.

Once the environmental conditions have been determined, the potential sources of uncertainty of each part of the instrument channel can be identified. For example, the process interface part is normally affected only by the process measurement and is not affected by equipment calibration or by other uncertainties. Furthermore, a cable in environmental condition B will not suffer a significant loss of measurement signal caused by a fall in insulation resistance.

Figure 1 shows the main types of uncertainty of each part of a typical instrument channel. Each of the main types listed can be subdivided further into more detailed specific types and specific parts. The following does not contain all the information that is required.

Process measurement effects include:

a) the effect of vessel and impulse line temperature, see Annex A;

b) the effect of liquid density on flow measurement, see Annex B;

c) the effect of the pipe configuration on flow measurement, see Annex B;

d) the effect of pipe pressure loss and head, see Annex C.

Instrument uncertainties:

a) reference accuracy, see Clause 3 and 6.2.7;

b) temperature effect, see 6.2.3;

c) pressure effect, see 6.2.4;

d) drift, see 6.2.8;

e) variation of the module power supply, see 6.2.10;

f) digital signal conditioning, see 6.2.10;

g) confirmation of RTD accuracy, see Annex D;

h) accident environment effects, see 6.2.5;

i) calibration effects, see 6.2.7.

Others:

a) the effect of insulation resistance, see Annex E;

b) wiring effects.

Once the above information has been obtained, the allowable values of the uncertainties have to be given. Many factors influence the allowable value; for example the process measurement effects derived from the analysis of the manufacturer of the nuclear steam supply system, the manufacturer's product specification and test reports, or actual plant data. In addition, if data cannot be obtained, and if there are limiting conditions that have to be respected in order to ensure that the result of the calculation is valid, different assumptions have to be made. Once they have been clearly listed, these uncertainties should be classified as random, bias or abnormally distributed.

Once the instrument channel diagram has been completed and the allowable values of the instrument uncertainties are known, the uncertainty of the whole instrument channel, from the process through to the bistable, can be expressed by a mathematical expression of the input-output relationships between the individual modules. The total uncertainty of the whole instrument channel can then be determined. The combination of the individual uncertainties into the allowable value of the total module and/or of the instrument channel is discussed in 6.3.

Finally, once the instrument channel uncertainty and the analytical limit are known, the trip setpoint and the allowable value can be determined. That part is discussed in Clause 7.

A flow chart of the process of determining the setpoint is given in Figure 2.

6 Calculation of instrument channel uncertainty

6.1 Uncertainty formulae

Because measurement is imperfect, the true value, that is the defined value of the measurand, cannot be obtained by measurement; the value obtained by measurement is only an estimate of the measurand. It is therefore not possible to obtain the actual error by measurement either. The total error should be quantified only from the point of view of probability; for example, one probability formula used for a given measurement guarantees correctness within a certain population, while measuring so as to satisfy the requirements of a different population requires a different probability formula. In this standard the term uncertainty is used to express the possible distribution of the error.