Improving Safety Culture:

Recognizing the Underlying

Assumptions

Douglas M. Minnema, PhD, CHP

Technical Staff Member
Defense Nuclear Facilities Safety Board

The views expressed are solely those of the author and no official support
or endorsement of this summary by the Defense Nuclear Facilities Safety
Board is intended or should be inferred.
 Purpose
• To consider safety culture lessons learned from the
Davis-Besse Reactor Pressure Vessel Head
Corrosion Event of 2002
• To conduct a historical review of other significant
incidents involving organizational weaknesses
• To develop an event progression model, based on
similarities between these incidents, to evaluate the
underlying assumptions in safety culture

ISM Workshop, November 2007 2

 Background
• INPO defines safety culture as:
“An organization’s values and behaviors –
modeled by its leaders and internalized by its
members – that serve to make nuclear safety
the overriding priority.”

• Note that there are four underlying assumptions

in this definition:
– Quality leadership - Prioritization
– Adequate span of control - Oversight
ISM Workshop, November 2007 3
 Background (con’t)
• INPO explains organizational culture as:
“The basic assumptions that have worked well
enough to be considered valid are taught to new
members of the organization as the correct way
to perceive, think, act, and feel. Culture is the
sum total of a group’s learning.”

• “Real progress on safety can be made by

understanding how people create safety, and
understanding how … safety can break down in
resource limited systems.” - Sydney Dekker
ISM Workshop, November 2007 4
 Consider
• If culture is the sum of an organization’s
learning, then a stable organization probably
unwittingly compensates for its own weaknesses
• Stability is maintained by peer pressure
“training” of new workers to the in-place culture,
unless external influences upset status quo
• Weaknesses often undetected until upset occurs

What happens when an organization

is stressed by external influences?
ISM Workshop, November 2007 5
 External Influences
External Influences are conditions that are outside
of organization’s span of control; the organization
can respond to them but cannot eliminate them.
• Market or economic conditions
• Industry deregulation
• Externally controlled mission, budget, or schedule
• Public opinion of industry or its products
• Competing externally-driven priorities (e.g.,
security, environmental protection)
ISM Workshop, November 2007 6
 At Davis-Besse
Over the several years before corrosion was discovered:
1. External market conditions drove managers to defer
maintenance; undertake merger; delay upgrades
2. Workforce responded by reducing rigor; chasing symptoms
not causes; raising informal issues threshold
3. Minimalist compliance approach evolved in organization,
failed to recognize significance of emerging issues
4. Oversight processes were not effective, if conducted, in
detecting changing organizational behaviors
5. Safety performance suffered general decline until
corrosion was discovered & regulator intervened

ISM Workshop, November 2007 7

 Event Progression Model
• This sequence suggests a general model:

External Influences Managers respond Workers respond by

Create pressure with org changes changing behaviors

Yes Yes

Does oversight Safety Performance Significance to

No No Safety recognized?
recognize decline? Declines

Significant Event Occurs

ISM Workshop, November 2007 8

 Is the Model Valid?
A review of major accidents and near-misses where
organizational weaknesses were causal factors demonstrated
the same event progression in 20 events:
Apollo I USS Forrestal TMI-2 NPP Iran Hostage
Fire, 1967 Fire, 1967 Meltdown, 1979 Rescue, 1980
UC Leak at UC Leak at Davis-Besse NPP Challenger
Bhopal India, 1984 Institute WV, 1985 LORF, 1985 Explosion, 1986
Chernobyl NPP Millstone NPP Maine Yankee Valujet #592
Explosion, 1986 Shutdown, 1996 NPP, 1996 Crash, 1996
CA Electrical Davis-Besse NPP Columbia NE Electrical
Brownouts, 2000 RPVH, 2002 Breakup, 2003 Outage, 2003
Los Alamos Lab BP Texas City BP Prudhoe Bay USAF Weapons
Shutdown, 2004 Explosion, 2005 Leak, 2006 Handling, 2007

ISM Workshop, November 2007 9

 Observations
• Not all accidents develop in this manner, but the
model does help understand how organizations
react to external influences
• “Good” leaders going in the wrong direction yields
same result as “bad” leaders with no direction
• Workers may follow managers’ lead, or they may
reject managers and rebel against organization;
either behavior leads to same results
• “Repeat performances” demonstrates the inherent
difficulties of establishing and sustaining a strong
culture within large organizations

ISM Workshop, November 2007 10

 Observations (con’t)
• External conditions strongly influence culture; while
external conditions cannot be eliminated, their
impacts on safety culture can be minimized
through understanding and proper planning
• The impacts of organizational changes on safety
culture should always be evaluated before the
change and then monitored after the change to
validate assumptions & conclusions

ISM Workshop, November 2007 11

 Observations (con’t)
• The model’s feedback loops of significance
recognition and oversight are how a leader
determines the health of the organization’s
culture before an event occurs; they should
never be neglected or ignored
• The motivation to hold nuclear safety as the
“overriding priority” requires leadership; an
organization will follow its manager’s behavioral
lead above the words; motivation cannot be
‘engineered’ into an organization
“People do what people see.” John C. Maxwell
ISM Workshop, November 2007 12
 Cultural Indicators
The value of a model is that one can identify
metrics based on it to monitor the culture:
– Changes in economic or market conditions
– Changes in production/service demands
– Changes in competing priorities
– Changes in safety versus non-safety resource
allocations
– Changes in workers’ satisfaction with
management
– Changes in turnover/retirement rates
ISM Workshop, November 2007 13
 Cultural Indicators (con’t)
– Changes in rate & nature of employee
concerns or dissenting opinions
– Changes in reportable events (up or down)
– Changes in rate of deferred maintenance
– Changes in rate of deferred/overdue training
– Changes in fraction of events involving
multiple programmatic breakdowns
– Changes in rate & nature of interactions with
regulators
ISM Workshop, November 2007 14
 Cultural Indicators (con’t)
– Changes in rate of overdue/delayed audits
– Changes in number or quality of findings
– Changes in turnover rates of auditors and
trainers
– Changes in ratio of externally- versus
internally-identified assessment/audit findings
– Changes in rate of overdue corrective actions
– Changes in ratio of self-disclosing versus self-
identified issues
ISM Workshop, November 2007 15
 Conclusions
• The underlying assumptions provide the keys to shaping
the organization’s behavior.
• Quality leadership cannot be engineered, but can be
developed; focus on creating a pool of future leaders.
• Influences outside the organization’s span of control can
have serious impacts on safety culture; understanding
those influences helps limit their impacts.
• The capacity to recognize the safety significance of
issues, and the ability to optimally prioritize resources in
response, require deliberate and constant attention
• The organization’s cultural health must be constantly
monitored through oversight, and early intervention into
identified concerns is always warranted.

ISM Workshop, November 2007 16

CLIMBING THE STEPS
TO AN EFFECTIVE
SAFETY CULTURE

SHARED DESIRE
FOR EXCELLENCE

CHECKS & BALANCES

EMPOWERED WORKERS

COMMITTED LEADERSHIP

BALANCED PRIORITIES & RESOURCES

TOOLS: VPP, QA, TRAINING, HPI, STANDARDS

FOUNDATION: INTEGRATED SAFETY MANAGEMENT

 •Selecting personnel based on values
Safety •Planning for succession
•Building teams
Culture •Recognizing contributions
•Sharing narratives and stories
•Teaching & mentoring on core values
•Developing future leaders

Shared Values •Cultivating a questioning attitude

And Beliefs •Establishing procedures and practices

•Analyzing and controlling work
•Training & qualifying workers and managers
•Ensuring continual competency
LEADERSHIP

Organizational

OVERSIGHT
•Determining the significance of issues
•Gathering feedback and improving
Practices •Learning from other’s lessons

•Identifying the standards & requirements

•Assuring a quality operation
Institutional •Finding and fixing issues
•Determining the basis for safety
Structures •Understanding the technical basis
•Strengthening institutional safety programs

Management •Applying Integrated Safety, Safeguards &

Security, and Environmental Management systems
Systems •Managing impacts of organizational change
•Balancing priorities during budget formulation
ISM Workshop, November 2007 18
 Conclusions (con’t)
• “Psychologists tell us that all humans have four
basic needs: the need to feel comfortable, the
need to be understood, the need to feel
welcome, and the need to feel important.”
– Mark Eppler
• Improving safety culture is all about satisfying
those four needs. To do so requires caring,
motivational leadership; the ability to determine
the significance of issues; and the capacity to
continually monitor the culture’s health.

ISM Workshop, November 2007 19