Development of a Communication Database of MCR Operators in - - PowerPoint PPT Presentation

Development of a Communication Database of MCR Operators in Emergency Situations in NPPs

Man Cheol Kim, Jinkyun Park, Wondea Jung Integrated Risk Assessment Center Korea Atomic Energy Research Institute

IEEE-HPRCT Conference

Korea Atomic Energy Research Institute 2

Outline

• 1. Introduction

1.1 Importance of correct communications

• 2. Necessities of a Communication Database

2.1 Results from investigations on human errors in NPPs 2.2 Need of communication-related human error data in HRA

• 3. Development of a Communication Database

3.1 Data Source 3.2 Extractions of verbal communications

• 4. Some Insights from Communication Analysis

4.1 Top-down communication type 4.2 Diversity in communication protocols 4.3 Diversity in sentence completeness

• 5. Conclusions

Korea Atomic Energy Research Institute 3

• 1. Introduction

1.1 Importance of correct communications

ν

Taking off without cabin crews

ν

July 30, 2006

ν

Taking off an airport at 1:56pm after checking the boarding of crews

ν

Ground crew confirmed that all crews are on board, understanding crews as flight crews

ν

Pilots took off the plane, understanding as all flight crews and cabin crews are on board

ν

After finding out that there are no cabin crew, the airplane returned to the airport after 30 minutes

ν

Communication error

ν

Ground crew : crew = flight crews

ν

Pilots : crew = flight crews + cabin crews (Source: D. Y. Song, Hangook Daily News)

Sorry, I forgot.

Korea Atomic Energy Research Institute 4

• 1. Introduction

1.1 Importance of correct communications

ν

Examples of communication errors in NPP operation

ν

Reactor trip in an abnormal situation

ν

During the process of increasing the reactor power, a reactor trip occurs due to a communication failure between the RO and the TO.

ν

TO said : “The feedwater supply pressure is normal.”

ν

RO misunderstood as “The feedwater line pressure is normal.”

ν

Reactor trip occurs due to the condition “Reactor power above 9% + feedwater line pressure low”

ν

Inappropriate reactor cooldown during an abnormal situation

ν

Due the management of an abnormal situation, an inappropriate reactor cooldown occurs due to a communication failure between the SRO and the RO.

ν

SRO said “Stop steaming.”

ν

RO switched the high pressure steam dump into the manual mode and closed the valve completely.

ν

Due to the increase in the steam generator pressure, the atmosphere steam dump valve opened at 1020 psig, and the reactor was cooled down to 548F inappropriately.

Korea Atomic Energy Research Institute 5

• 2. Necessities of a Communication Database

2.1 Results from investigations on human errors in NPPs

ν

Analysis results of large-scale investigations

ν

Germany

ν

Analysis of 232 incidents including human errors [Strater, 2004]

ν

165 cases in BWR plant, 55 cases in PWR plants, and later added 12 more cases for communication analysis

ν

In about 10% of the incidents induced by human errors, communication errors worked as an important contributor

ν

Operational aspects such as workload and situational stress affected the communication, and the failure in communication affected the operational aspects (therefore, they cannot be considered independently.)

ν

Japan

ν

Analysis of 193 incidents including human errors [Hirotsu, 2001]

ν

Total number of the incidents were 885, and therefore the human errors take the 21.8% of importance.

ν

About 13% of the incidents including human errors were caused by written communication, and about 5% of them are cause by verbal communication

Korea Atomic Energy Research Institute 6

• 2. Necessities of a Communication Database

2.2 Need of communication-related human error data in HRA

ν

Providing base data for estimating communication-related PSFs [NUREG/CR-6903]

PSF comparison between Good Practices, SPAR-H, and HERA

Korea Atomic Energy Research Institute 7

• 3. Development of a Communication Database

3.1 Data Source

ν

Three methods for data gathering related to communication analysis [Fukuda, 2004]

ν

Incident reports

ν

Communication errors can be found in abnormal or emergency situations.

ν

In most cases, communication logs are not available

ν

Operational field studies

ν

Inefficient because the abnormal or emergency situations do not occur frequently

ν

Training in simulated situations using a full-scope simulator

ν

Provides a lot of experimental data

ν

Concerns about the applicability of the data gathered in simulated situations to real situations

→

It will be meaningful to use the data gathered in simulated situations until we can gather sufficient data in real situations.

Korea Atomic Energy Research Institute 8

• 3. Development of a Communication Database

3.1 Data Source

ν

What to observe?

ν

NPP operators’ management of simulated emergency situations in a full-scope simulator for OPR-1000 (KSNP) NPPs

ν

Number of observations

ν

10 MCR operator teams

ν

Simulated emergency situations

ν

Loss of coolant accidents (LOCAs)

ν

Steam generator tube rupture (SGTR) accidents

ν

Loss of all feedwater (LOAF) accidents

ν

Excessive steam dump event (ESDE) accidents

Korea Atomic Energy Research Institute 9

• 3. Development of a Communication Database

3.1 Data Source

Operator Team Accident Scenario Team 1 SGTR Team 2 SGTR Team 3 LOCA Team 4 ESDE Team 5 SGTR Team 6 LOCA Team 7 ESDE Team 8 LOAF Team 9 SGTR Team 10 ESDE Scenario LOCA SGTR LOAF ESDE Number

• f use

2 4 1 3 Used simulated accident scenarios and a summary of the number of use

Korea Atomic Energy Research Institute 10

• 3. Development of a Communication Database

3.2 Extractions of verbal communications

ν

An example communication data from the aviation industry

Time Source Content 42:20 HOT-2 Have a nice flight 42:30 CAM [sound of increasing engine noise] 42:31 HOT-1 EPR select 42:32 HOT-2 EPR 42:38 HOT-2 Power’s set 42:40 HOT-1 Okay checked 42:45 HOT-2 Eighty knots 42:46 HOT-1 Checked 42:48 HOT-1 My airspeed indicator’s not working 42:50 HOT-2 Yes 42:51 HOT-2 Yours is not working

HOT-1 = captain HOT-2 = co-pilot CAM = cockpit area microphone [ ] = editorial insertion

Korea Atomic Energy Research Institute 11

• 3. Development of a Communication Database

3.2 Extractions of verbal communications

ν

A part of communication database developed by KAERI

Beginning time End time Duration Speaker Content 00:00.0 00:03.2 00:03.2 RO I am going to trip the reactor. Three, two, one, trip. 00:03.3 (Reactor trip) 00:07.5 00:08.7 00:01.1 TO The turbine is tripped. 00:08.8 00:13.6 00:04.8 SRO Because the plant is shut down, we are going to perform SPTAs. 00:16.4 00:18.7 00:02.3 SRO RO, reactor power? 00:19.1 00:24.6 00:05.4 RO Reactor power... All control rods are inserted, reactor power is decreasing, and the startup rate maintains the minus value. 00:25.0 00:27.0 00:02.1 SRO Is there any control rod that cannot be inserted? 00:27.1 00:27.8 00:00.7 RO No. 00:28.6 00:30.7 00:02.1 SRO Next, turbine trip? 00:32.4 00:33.9 00:01.5 TO The turbine is tripped. 00:34.7 00:36.9 00:02.3 SRO Next, opening of the power circuit breakers of the main transformer? 00:37.0 00:38.2 00:01.2 EO Yes, they are opened.

Korea Atomic Energy Research Institute 12

• 4. Some Insights from Communication Analysis

ν

What is a good communication?

An imaginary relation between the communication and the performance of NPP operators

Korea Atomic Energy Research Institute 13

• 4. Some Insights from Communication Analysis

4.1 Top-down communication type

ν

Four communication types of NPP operators [Ujita et al., 1992]

ν

Top Down

ν

Top down instruction and bottom up information

ν

Bottom Up

ν

Instruction and information among bottom, bottom up information, and top down approval

ν

Tight Coupling

ν

Both directed instruction and information

ν

Loose Coupling

ν

Poor instruction and information

ν

Relation between communication and performance of NPP operators

ν

Good performance was observed with the tight coupling communication type.

ν

Observed communication type in simulated emergency operations

ν

Top down communication type was usually observed, due to the performance of emergency operating procedures (EOPs).

Korea Atomic Energy Research Institute 14

• 4. Some Insights from Communication Analysis

4.2 Diversity in communication protocols

ν

Use of different communication protocols

ν

An example of SRO’s different communication protocols

ν

SRO 1

ν

Is the reactor shut down ?

ν

SRO 2

ν

Is the reactivity control function satisfied ?

ν

SRO 3

ν

Is the reactor power decreasing ?

ν

Are all control rods at their bottom position ?

ν

Is the startup rate minus ?

ν

An example of RO’s different communication protocols

ν

RO 1

ν

The reactor is properly shut down.

ν

RO 2

ν

The reactor power is decreasing, all control rods are at their bottom position, and the startup rate is minus.

Korea Atomic Energy Research Institute 15

• 4. Some Insights from Communication Analysis

4.2 Diversity in communication protocols

ν

Development of a standard communication protocol

A part of a developed standard communication protocol

Korea Atomic Energy Research Institute 16

• 4. Some Insights from Communication Analysis

4.2 Diversity in communication protocols

A comparison of workloads evaluated by using NASA-TLX

Korea Atomic Energy Research Institute 17

• 4. Some Insights from Communication Analysis

4.2 Diversity in communication protocols

A comparison of some characteristics related to the safety of NPPs

Korea Atomic Energy Research Institute 18

• 4. Some Insights from Communication Analysis

4.3 Diversity in sentence completeness

A, 91 A, 90 A, 86 B, 8 B, 10 B, 11 C, 1 C, 0 C, 2 10 20 30 40 50 60 70 80 90 100 Subject Object Verb A, 87 A, 80 A, 65 B, 13 B, 16 B, 28 C, 0 C, 3 C, 8 10 20 30 40 50 60 70 80 90 100 Subject Object Verb A, 89 A, 80 A, 75 B, 11 B, 16 B, 12 C, 0 C, 4 C, 13 10 20 30 40 50 60 70 80 90 100 Subject Object Verb A, 60 A, 85 A, 86 B, 40 B, 11 B, 10 C, 0 C, 4 C, 4 10 20 30 40 50 60 70 80 90 100 Subject Object Verb

(a) ESDE – Team 1 (b) ESDE – Team 2 (c) LOAF – Team 1 (d) LOAF – Team 2

A comparison of the sentence completeness between two NPP operator teams

Korea Atomic Energy Research Institute 19

• 5. Conclusions

ν

Importance of good communications in NPP operation

ν

Communication errors have been found in the real operation of NPPs

ν

Reactor trip due to a communication failure between the RO and the TO

ν

Inappropriate reactor cooldown due to a communication failure between the SRO and the RO

ν

Large-scale investigations in human error found that communication errors are significant.

ν

In a German study, communication errors are identified to take about 10% of all human errors found in incidents.

ν

In a Japanese study, communication errors are identified to take about 18% of all human errors found in incidents.

ν

Development of a communication database

ν

Method

ν

Operator training in simulated emergency situations

ν

10 licensed operators teams of OPR-1000 (KSNP) NPPs were participated.

ν

Simulated accident scenarios consist of LOCA, SGTR, LOAF, and ESDE.

Korea Atomic Energy Research Institute 20

• 5. Conclusions

ν

Some insights from observations on NPP operators’ communications in simulated emergency situations

ν

Top-down type communications

ν

Diversity in communication protocols

ν

Diversity in sentence completeness