CCPS PERD Reliability API Presentation_r3

Report
PERD
A discussion with the API Subcommittee on
Instruments and Control Systems
23 April 2012
2
Your Presenters

Gene Meyer
 Dow Chemical, 28 years
 Reliability Engineering for the last 12 years.

Hal Thomas (on the phone)
 Exida (formerly Air Products)
 Process Safety
 Original chair of PERD
3
Alignment

AICHE
 CCPS, Center for Chemical Process Safety
 PERD, Process Equipment Reliability Database
• Participants
» Members and Users
 A subscription-based project with CCPS.
4
Agenda
•
Introduction
•
Alignment
•
Organization
•
Goal
•
The solution
•
Details
•
Results
•
Discussion
5
Reliability Situation




Industry-specific reliability data is not available.
Quality company data is scarce.
Performance-based regulations enacted.
CPI and HPI capital intensity forces greater asset
management (i.e. higher process reliability)
6
Process Safety Situation
 Performance-based regulations enacted.
 Lack of quality data reduces confidence & jeopardizes industry
credibility with respect to quantitative analysis results
 Performance-based processes require quality management systems
that are auditable and defensible
 Quality Assurance requires data to validate performance assumptions
7
Current Reliability Modeling Situation
8
The Vision
Readily available Maintenance and Reliability Data that:
 Is generated from equipment like yours,
 Is automatically loaded into a database,
 Represents industry experience,
 Is validated for accuracy,
 Is statistically evaluated by documented methods,
 Is benchmarked versus your industry
 Is available for your use in:

Simulation

Process validation

Process design and troubleshooting

Reliability Growth

Quantitative Risk Assessment
9
The Solution
PERD
 Process Equipment Reliability Database
 A database that works with your computerized maintenance
management system to upload equipment performance data, organize,
validate, calculate, benchmark and make available for use with clear
statistical guidance.
 The database that uses company and industry equipment
performance data evaluated against industry standards.
PERD is:
A member-led project developed by CCPS to provide
equipment reliability data with the following mission :
The PERD Mission
Operation of an Equipment Reliability Database, Making
Available High Quality, Valid, and Useful Data to the HPI
and CPI Enabling Analyses to Support Availability,
Reliability, and Equipment Design Improvements,
Maintenance Strategies, and Life Cycle Cost Determination
11
PERD Provides:
 Practical approach and Consistent work processes for data
collection

Sound theoretical foundation based on engineering fundamentals
 Efficient software tools for data submission

S/W dedicated for use by participating companies
 Quality assurance for data analysis

Identification and analysis of outliers

Defensible and auditable results
12
PERD History
 During development of CCPS Chemical Process Quantitative Risk
Analysis book, committee concluded a separate guideline book was
needed to address data
 1989 - The result was Guidelines for Process Equipment Reliability
Data with Data Tables
 Following publication, Hal Thomas and Tom Carmody attended an
OREDA meeting and concluded data mining was too expensive, but
data harvesting if implemented properly was sustainable, hence the
PERD initiative was born.
 1998 – Guidelines for Improving Plant Reliability Through Data
Collection and Analysis published establishing the basic
foundational approach
 Today – Work processes established and documented, rolling out
2nd generation PERD software, continuing to develop equipment
taxonomies & analyze equipment data
13
Fundamental Concepts
 Develop & document equipment boundary diagram
 Perform rigorous “Functional Analysis” to determine and
document all of the fundamental failure modes that
represent complete or degraded failures as well as
documenting incipient conditions for an equipment type
 Reference paper by Rausand and Oin in their critique of OREDA
database
14
Fundamental Concepts Continued
 Engineer information systems to
 Support data acquisition using a experimental design mentality,
facilitating immediate data analysis
 Input only facts, i.e. do not require interpretation on part of
mechanic or technician
• Make use of standard pick lists to extent possible
 Create event input forms that allow inference of failure modes by
asking the question, ‘What data is it that would allow us to
conclude a particular failure mode has occurred.
15
Fundamental Concepts Continued
 Leverage existing information systems that are going to
exist anyway, improving data quality with no increase in
cost
 Design equipment design spec data fields to be PERD
compatible
 Design process demand, inspection and test record data fields
to be PERD compatible
163Feb00_Ins
urance
Fundamental Concepts - The Data Farm
Quality info for use in reliability
analysis, maintenance strategies,
risk analysis, etc.
Events occur
on a
continuous
basis
Electronically
extract
validated data
Periodic batch
transfer of
data
PERD Database
16
An Integrated Approach
CCPS PERD
Industry Offline Reliability
Analysis Database
Process Specs
In-Tools
Smart P&IDProcess Specs
Design Specs
Maintenance Module
(i.e. SAP, Maximo, etc.)
Corporate PERD
Purchase Order
Population Data
Installation
Process Specs
Facility PERD
Commissioning
Maintenance
Activities
Operating Group
Information System(s)
Plant Operation
Process Events
The Harvest
CCPS Industry
Database
Inventory Data
ID Number
Tag Number
Company 1
Database
Company 2
Database
Company n
Database
Plant 1
Data
Plant 2
Data
Plant n
Data
Inventory Data
Event Data
Event Data
Maintenance
Inspections
Proof Testing
Incidents
Etc.
Inventory Data
Event Data
19
Initiative Focus
 Sound theoretical foundation
 Practical application to plant and equipment, including
software tools
 Fundamentally sound engineering information available to
companies to “Engineer” their information management
systems
 Industry database built upon the platform above
 Capitalize on opportunities
 Company databases adapt to the platform above
Fundamental Taxonomy Relationships
Inventory Record
Application Data Fields
Tag Number
Operating
Conditions
Device Data Fields
Equip ID
- Equip Descriptors
Events
Operating
Mode
Equipment Function(s)
Relationship Legend
Many to Many
Failure Mode(s)
One to Many
Many to One
Failure
Mechanism(s)
Failure
Cause(s)
One to One
The same is true for
Reliability &
simulation
22
Built Upon a Sound Technical Foundation
 PERD Has Adopted a Data Tier Structure That:
 Eliminates Data Intimidation
 Allows more rapid expansion of partial Equipment taxonomies
and associated software development
 Basic Tier Concept
 Tier 1 – Whatever minimal data might exist – Gets people
started; generates a useful event record. Pass/fail
 Tier 2 – Begins to provide event specifics; population filtering
and failure classification
 Tier 3 – Top of the “PERD ladder of success”; failure modes
23
Equipment Groups & Systems
Equipment
Groups
Electrical
Fixed Equipment
Instrumentation
Equipment
Systems
 Circuit Breaker
 Contactors
 Motor
 Relays
 Transformer
 Cylinder
 Heat Exchanger
 Piping Circuit
 Pressure Vessel
 Atm Storage Vessels
 Analyzer
 Annunciator
 Fire & Gas
Detection
 Loop
 Sensing Element
 Switch
 Transmitter
 Signal Conditioning
Unit
 Control Logic Unit
Expandable
Pressure Relieving
Devices
 Reclosable pressure
Relief Devices
 Non-reclosable
Pressure Relief
Devices
 Hydraulic
Accumulator
 Liquid Seal
 Open Vent
Rotating
Machinery
 Agitator
 Compressor
 Pump
 Turbine
Valves
 Check Valve
 Hand Valve
 Self-Actuated
 Solenoid Valve
 Remote Actuated
Inventory Data Tier Concept

Tier 1
 Location Address (e.g. Plant, tag, etc.)
 Equipment Group (e.g. Instrumentation)
 Equipment System (e.g. Transmitter)
*Equip Group
PL_Equip_Grp
*Equip System
PL_Equip_Sys
*Tag Number
*Equip ID
Comments
25
Equipment
Type & Subtype
Example
Equipment
Group
Pressure Relieving
Devices
Equipment System
Reclosable pressure
relief devices
Non-reclosable pressure
relief devices
Hydraulic accumulator
Liquid seal
Open vent
Equipment Type
Pressure Relief Valve
(Spring Loaded)
Pilot Operated Relief
Valve
Conservation Vent
Self-Closing Manhole
Cover
Rupture Disk
Explosion Vent
Blow-off Man-way
Rupture Pin
Buckling Pin
Weak Gauge Hatch
Fusible plug
Weak shell to roof seam
Equipment Subtype
PL_REL_SubType
 Conventional
 Balanced Bellows
 Balanced Bellows With
Auxiliary Balanced Piston
Inventory Data Tier Concept

Tier 2
 All tier 1 data plus following data
user chooses to enter
 Equipment Type (e.g. Differential
Pressure)
 Equipment Subtype (e.g. Level)
Transmitter Type
Transmitter
Subtype
Manufacturer
PL_Equip_Type
PL_Equip_SubType
PL_Mfr_Equip
Model Number
 Manufacturer
 Model Number
 Output Signal or Set Point (As
applicable to equipment)
Output Signal
Set Point
PL_Equip_Sig_Out
PL_Equip_Set_Point
Whichever field
is applicable
Inventory Data Tier Concept
 Tier 3
 All tier 1 data
 Equipment type
 Equipment subtype
 Plus additional tier 2 and / or 3 data user chooses to enter

Possible to achieve level of a complete equipment
specification
Event Data Tier Concept

All Events require
• Linkage to tracked equipment
• Event date (and time if applicable)

Tier 1
• Data showing a failure has occurred

Tier 2
• Data sufficient to infer whether failure was “Dangerous” or
“Safe”

Tier 3
• Data sufficient to infer true failure modes
User decides what event data to track!! Event data tracked
determines what level of inference is achievable.
29
Relief Valve Proof Test Example
30
Getting Started
 Upon Joining as a Full Participant
 PERD provides the software & initial help to companies
loading the software onto their company server
 PERD provides initial help and guidance allowing you to
contribute whatever data you have, enabling companies to
quickly get involved and begin accruing benefits
 Company data contribution means access to aggregated
anonymous industry data and a path to continuous
improvement
31
Helping Hand Work Process – Getting Started
Ad Hoc Data Acquisition & Analysis
Existing CMMS or
other Info Mgmt
System
Company
downloads
data
Company
submits data to
CCPS PERD
PERD
Guidance
Helping Hand Work Process
PERD
Performs
Data Map
PERD translates
data &
populates PERD
compatible files
PERD Imports
Data & Manages
Dataset
PERD Aggregates
Data
(Multiple Data
Subscribers)
PERD Anonymizes
Data
Data Analysis
32
Immediate Benefits
 Access to fundamental technical information that documents specific
data user needs to:
 Filter inventory data for populations of interest
 Infer failure modes of interest
 Access to work process knowledge needed to achieve company
goals
 Facilitates moving to data farming (low cost) work process rather
than data mining (high cost) activity
 Enables maximization of benefits from enterprise systems (eg SAP,
Maximo, Matrikon, Meridian, etc.)
 Provides pre-engineered solutions that can be implemented cost
effectively
 Best available info transfer between engineering and
programmers
33
On-Going Work Process
On Going Data Acquisition & Analysis
Existing CMMS or
other Info Mgmt
System
Company
downloads data in
compatible
format
Company
Documents Quality
Plan
Company submit
data to CCPS PERD
per quality plan
PERD import data
into industry
database
Quality Process
PERD Aggregates
data from
(Multiple Data
Subscribers)
PERD makes data
anonymous
Data
Data analysis
analysis
34
Long Term Benefits
 Data Harvesting per Defined Quality Plans Provide:
 Operating company database capable of achieving
proven in use data for safety and reliability analysis
 Data Contribution to PERD Facilitates
 Operation of industry database
 Industry benchmarking
 Statistical analysis by true experts due to best data
available to support their research
 Data available for improved reliability and risk analysis
 On going continuous improvement via reliability
growth
35
Why Join PERD?
PERD enables user companies to leverage existing IT and
MI systems to collect and submit data resulting in:
 Improved Reliability
 Industry benchmarking
 Identification of “low hanging fruit” for immediate impact
 Minimization of unforeseen losses
 Increased Effectiveness for

Information Systems Technology

Reliability & Quantitative Risk Assessments

Regulatory Compliance
For a small financial investment and minimal data.
36
You would join…
Full Participants
o Air Products
Subscribers & Volunteers
o Exida
o Chevron
o Rosemount
o DNV
o Savannah River Site
o Dow Chemical
o DuPont
o FM Global
o Sis-Tech
o TMC
37
Getting Started.
 Join as a Full Participant
 Benefits of Full Participant Status:
 PERD provides the software & initial help to companies
loading the software onto their company server
 PERD provides initial help and guidance allowing you to
contribute whatever data you have, enabling companies to
quickly get involved and begin accruing benefits
 Company data contribution means access to aggregated
anonymous industry data and a path to continuous
improvement
 Networking with industry experts involved with the PERD
project.
38
In Summary
 PERD provides you a roadmap to improved performance
whereby you can
 Improve Information Systems Effectiveness
 Create Value
 Minimize unforeseen losses
 Improve Regulatory Compliance Effectiveness
Enroll Now!!
39
For More Information Contact



Dave Belonger – CCPS PERD Staff Consultant
609-654-4914
[email protected]




Gene Meyer – Chairperson
Dow Chemical
(989) 638-4064
[email protected]




Hal Thomas – Technical Consultant
exida
610-481-9681
[email protected]
40

similar documents