ROOT CAUSE FAILURE ANALYSIS

OF
EOG-OILBIRD FIREWATER PUMP LOW DISCHARGE PRESSURE

Jameel Babooram
TOSL Engineering Ltd
11/9/2018
Page | 1
Abstract
An engineering study was performed to determine the Root Cause of low discharge pressure on a
multistage vertical firewater pump (FWP) for EOG-Oilbird facility.

The Aim of this study was to determine the possible cause(s) of low pump discharge pressure. To
fulfil the Aim, the following objectives were undertaken:

1 Perform Positive Material Identification (PMI) on the pump column pipes.

2 Perform visual inspection of the pump assembly.
3 Perform a Root Cause Failure Analysis (RCFA) to determine the root cause(s) of low pump
discharge pressure.
3 Develop an action plan to mitigate its reoccurrence.
Positive Material Identification revealed that the flange sections of the column pipes are 316
Stainless Steel however, the pipe section of the columns presently in use is Low Alloy Carbon
Steel and not A 106 CS as specified by the reference drawing. Visual inspection also revealed two
column pipe sections contained holes and the pump impeller assembly contained moderate build-
up of biofouling barnacles.

The RCFA study revealed the following as the root causes of low pump discharge pressure; lack
of awareness of the effects of damaged protective coating, Cathodic Protection System inefficient
for a mean time to failure of five years and lack of Biocide Dosing.

Control actions required to mitigate the low pump discharge pressure include; developing a
Standard Operating Procedure (SOP) for handling of the pipe columns in order to minimize the
damage to the protective coating, review & re-design the present Cathodic Protection System to
ensure a minimum Mean Time to Failure (MTTF) of 5 years if Low Alloy Carbon Steel is to be
used as the column pipe or upgrade the existing column pipes to 316 Stainless Steel, and develop
and implement a biocide dosing program for the FWP column pipe and pump sections.
Benchmarking was also performed with EOG-Toucan FWP column pipes and PMI revealed that
both the column flange and pipe sections are 316 Stainless Steel. The use of 316 Stainless Steel
throughout the column pipes eliminate the possibility of galvanic corrosion due to bimetallic
coupling of different materials. It also eliminates the need for cathodic protection which decreases
the preventive maintenance time and cost and increases the Mean Time to Failure of the FWP.
Therefore, based on the aforementioned advantages it is highly recommended that EOG Resources
Trinidad Limited consider replacing the existing column pipes with 316 Stainless Steel.

Page | 2
Table of Contents
1.0 Introduction ............................................................................................................................................ 5
2.0 Aim and Objectives ................................................................................................................................. 5
2.1 Aim ......................................................................................................................................................... 5
2.2 Objectives ............................................................................................................................................... 5
3.0 Literature Review .................................................................................................................................... 6
3.1 Root Cause Failure Analysis ................................................................................................................... 6
3.2 Corrosion................................................................................................................................................. 7
3.2.1 Galvanic Corrosion ....................................................................................................................... 7
3.3 Firewater Pump Specification ................................................................................................................. 8
3.3.1 Pump Specifications ..................................................................................................................... 8
3.3.2 Column Pipe Material Specification ............................................................................................. 8
4.0 Methodology........................................................................................................................................... 9
5 Root Cause Failure Analysis ..................................................................................................................... 11
5.1 Problem Definition ................................................................................................................................ 11
5.2 Brainstorming........................................................................................................................................ 12
5.2.1 Severity Ranking ......................................................................................................................... 13
5.3 Prioritization.......................................................................................................................................... 14
5.4 Cause and Effect Diagram ..................................................................................................................... 15
5.5 Fault Tree Analysis – Leak in the discharge piping system ................................................................... 16
5.5.1 Fault Tree Analysis – External Corrosion – Localized Corrosion ........................................................ 17
5.5.2 Fault Tree Analysis – External Corrosion – Galvanic Corrosion ......................................................... 18
5.5.3 Fault Tree Analysis – External Corrosion – Wrong Material Type ..................................................... 19
5.6 Fault Tree Analysis – Impellers partially clogged .................................................................................. 20
5.7 Summary of Findings............................................................................................................................. 21
6 Action Plan – Recommendations ............................................................................................................. 22
6.1 Action Plan – Localized Corrosion ......................................................................................................... 22
6.2 Action Plan – Galvanic Corrosion – Cathodic Protection System Inefficient ........................................ 22
6.3 Action Plan – Galvanic Corrosion – Wrong Material Type ................................................................... 23
6.4 Action Plan – Impellers partially clogged – Lack of Biocide Dosing ...................................................... 24
Appendix 1 – ACAD FILE:833602-PUMP-GA.DWG ...................................................................................... 25
Appendix 2 – TOSL-PO4656/PMI-001 ......................................................................................................... 26
Appendix 3 – Build-up of biofouling barnacles on impeller........................................................................ 28

Page | 3
Appendix 4 – Minimum build-up of biofouling barnacles on the suction strainer ..................................... 29
Appendix 5 – Proper sealing of column pipe flanges.................................................................................. 30
Appendix 6 – External Corrosion leading to holes in two pipe column sections........................................ 31
Appendix 7 – Impeller #1 with excessive clearance at the bore area ........................................................ 32
Appendix 8 – Anodic Index of Metals ......................................................................................................... 33
Appendix 9 – TOSL-PO4656/18/PMI-008 Toucan FWP .............................................................................. 34

Page | 4
1.0 Introduction

EOG Resources Trinidad engaged TOSL Engineering Services Ltd to perform an engineering
study to determine the Root Cause of low discharge pressure on a multistage vertical firewater
pump (FWP) for its Oilbird facility.

2.0 Aim and Objectives

2.1 Aim

The Aim of this study is to determine the possible cause(s) of low discharge pressure on a
multistage vertical firewater pump (FWP) for its Oilbird facility.

2.2 Objectives

To fulfil the Aim of this study the following objectives were undertaken:

1 Perform positive material identification on the pump column pipes.

2 Perform visual inspection of the pump assembly.

3 Perform RCFA to determine the root cause(s) of low discharge pressure.

3 Develop an action plan to mitigate its reoccurrence.

Page | 5
3.0 Literature Review

3.1 Root Cause Failure Analysis

Root Cause Failure Analysis (RCFA) is a widely used method in industries to determine the
cause(s) of failure and mitigate recurrence. Therefore, it is a failure analysis methodology focus
on detecting the underlying causes of failure and control similar failures in future by identifying
and eliminating those causes (York D, 2014).

The main goal of RCFA is to avoid and control recurrence of failures by eliminating root causes.
It seeks to identify the origin of a problem using specific steps and tools to find the root of the
problem in order to:

I. Determine what happened.

II. Determine why it happened.
III. Determine how to reduce the likelihood that of it reoccurring.

RCFA assumes that the system under the investigation and the events are interrelated. Therefore,
an action in one area triggers an action in another area. By methodically tracing actions, one can
discover where the problem started and how to minimize or eliminate its reoccurrence. Causes to
problems fall into three basic type of categories:

Physical Cause – Tangible, material items failed in some way.

Human Cause – Human error leads to physical causes.

Organizational Causes – A system, process, or policy that people use to make decisions.

RCFA methodology looks at all three types of causes and involves investigating the patterns of
negative effects, finding hidden flaws in systems and discovering specific actions that contribute
to the problems. This often means that RCFA reveals more than one root cause.

Page | 6
3.2 Corrosion

According to ISO 8044-1986 corrosion is defined as “Physicochemical interaction between a metal

and its environment which results in changes in the properties of the metal and may often lead to
impairment of the function of the metal, the environment, or the technical system of which these
form a part”. Hence, it is the gradual destruction of materials (usually metals) by chemical and/or
electrochemical reaction with their environment.

3.2.1 Galvanic Corrosion

Galvanic corrosion (also called 'dissimilar metal corrosion') refers to corrosion damage induced
when two dissimilar materials are coupled in a corrosive electrolyte. It occurs when two (or more)
dissimilar metals are brought into electrical contact under water. When a galvanic couple forms,
one of the metals in the couple becomes the anode and corrodes faster than it would all by itself,
while the other becomes the cathode and corrodes slower than it would alone (Nace-International,
2018).

Because different metals have different natural potentials in a liquid, a current will flow from the
anode (more electronegative) metal to the cathode (more electropositive), which will result in
increased corrosion at the anode.

A dissimilar (bi-metallic) couple is similar to a battery, consisting of two dissimilar metals

immersed in an electrolyte solution. An electric current (flow of electrons) is generated when the
two electrodes are connected by an external, conductive path. Figure 1 below highlights this
concept.

Figure 1: Galvanic Cell

Page | 7
There are four elements necessary for corrosion to occur in a galvanic cell:

 Anode - The electrode where galvanic reaction(s) generate electrons - negative ions are
discharged and positive ions are formed. Corrosion occurs at the anode.

 Cathode - The electrode that receives electrons - positive ions are discharged, negative ions
are formed. The cathode is protected from corrosion.

 Electrolyte–The conductor through which current is carried. Electrolytes include aqueous

solutions or other liquids.

 Return Current Path - The metallic pathway connecting the anode to the cathode. It is
often the underlying metal substrate.

3.3 Firewater Pump Specifications

3.3.1 Pump Specification

Goulds 14RJHC / 5 Stages

Reference – ACAD FILE: 833602-PUMP-GA.DWG

3.3.2 Column Pipe Material Specification

Flange

316 stainless steel – Reference ACAD FILE: 833602 –PUMP-GA.DWG

Pipe

A 106 CS Pipe – Reference ACAD FILE: 833602 –PUMP-GA.DWG

(See appendix 1 for further details)

Page | 8
4.0 Methodology

The methodology utilized for this study involved performing an initial investigation to classify and
define the problem. Once this was completed, a brainstorming exercise was undertaken followed
by a cause and effect analysis and fault tree analysis to map the potential causes into specific
categories and determine the root cause(s) of the problem. Once the root cause(s) were identified
an action plan was created to address each cause.

1 Problem Definition -

The problem definition was answered by utilizing the 5W analysis:

 What is the Problem?

 When the problem occurred?
 Where the problem occurred?
 Who is the witness for the problem?
 Why the problem has happened?

2 Brainstorming

This technique was used to identify the possible causes of internal corrosion:

 Freewriting - Word Storm

 Researching
 Listing
 Mapping

3 Cause and Effect Analysis – Fishbone (Ishikawa) Diagram

This technique was used to map the possible causes of internal corrosion into the following
categories:

 Methods
 Machines (equipment)
 People (manpower)

Page | 9
  Materials
 Measurement
 Environment

4 Fault Tree Analysis

Fault Tree Analysis (FTA) was used as a deductive analysis approach for resolving the undesired
event (possible causes) into its root cause(s) through the use of logic diagrams.

5 Action Plan

An action plan was used to lists what steps must be taken in order to eliminate or minimize the
risk of reoccurrence of the problem. It was also used to formulate a timeline for when specific
tasks need to be completed and determine who is responsible. Figure 2 below highlights the RCFA
methodology utilized for this study.

Problem
Definition

Brainstorming

Cause and
Effect Analysis

Fault Tree
Analysis

Action Plan

Figure 2: RCFA Methodology

Page | 10
5 Root Cause Failure Analysis
5.1 Problem Definition

5.1.1 What is the Problem?

Low discharge pressure of EOG-Oilbird Fire Water Pump.

5.1.2 When the problem occurred?

During the last 18 months in service.

5.1.3 Where the problem occurred?

EOG-Oilbird facility.

5.1.4 Who is the witness for the problem?

EOG Resources Trinidad personnel.

5.1.5 Why the problem has happened?

To be determined through the use of a RCFA methodology. However, based on visual inspections
holes were found in two sections of pump column pipes.

Page | 11
 5.2 Brainstorming

Brainstorming was conducted by a team of Engineers and Senior Maintenance and Inspection
Technicians to determine the possible reasons for low FWP discharge pressure. Figure 3 below
highlights the possible causes for low FWP discharge pressure.

Leak in the discharge

piping system*****

Impeller wear
rings worn*
Leaking column
joints **

Suction strainer
partially clogged Impeller
by biofouling loose**
barnicles**

Impeller partially
clogged by Damage
biofouling Impeller***
barnicles****

Figure 3: Possible causes of Gearbox Corrosion

Note: The number of asterisks (*) at the end of “a possible cause” in Figure 3 above indicates the
amount of times “the same possible cause” was highlighted by a different member of the RCFA
Team.

Page | 12
5.2.1 Severity Ranking
The possible causes identified in Figure 2 above were ranked based on the likelihood of occurrence and
highlighted in Table 1 below.

Likelihood
Asterisk
No. Possible Cause of Remarks
(*) Level
occurrence
Visual inspection revealed
Impeller partially moderate build-up of
1 4 clogged with Moderate biofouling barnacles on all
biofouling barnacles impellers (see appendix 3 for
further details).
Visual inspection revealed
Suction strainer is minimal build-up of biofouling
2 2 patially clogged by Very Low barnacles on suction strainer
biofouling barnicles (see appendix 4 for further
details).
Visual inspection revealed
Leaking column proper sealing of all column
3 2 Very Low
joints pipe flanges (see appendix 5
for further details).
Visual inspection revealed two
Leak in the
column pipe sections contained
4 5 discharge piping Very High
holes (See appendix 6 for
system
further details).
Visual inspections revealed
5 1 Wear rings worn Very Low wear rings in acceptable
condition.
One impeller showed
excessive clearance at the shaft
bore area however, piece of
6 2 Impeller Loose Very Low shim stock was found lodged
to make up the clearance
necessary for an ideal fit. (See
appendix 7 for further details).
Visual inspections revealed
7 3 Damage impeller Very Low impellers in acceptable
condition.

Table 1:Severity Ranking of Possible Causes

Page | 13
Asterisk Level - The number of asterisks (*) at the end of “a possible cause” above indicates the
amount of times “the same possible cause” was highlighted by s different member of the RCFA
Team.

Possible Cause – Reason for low discharge pressure of EOG – Oilbird FWP.

Likelihood of Occurrence – Very High (5), High (4), Moderate (3), Low (2), Very Low (1).

Remarks – Reason for the likelihood of occurrence.

5.3 Prioritization

The possible causes for pump low discharge pressure were prioritized based on likelihood of
occurrence and severity level, highlighted in Table 2 below.

Likelihood of Severity
No. Possible Cause
Occurrence Level
1 Leak in the discharge piping system Very High 5
2 Impeller partially clogged Moderate 3

3 Impeller loose Very Low 1

Suction strainer is patially clogged by
4 biofouling barnicles Very Low 1

5 Leaking column joints Very Low 1

6 Wear rings worn Very Low 1

7 Damage impeller Very Low 1

Table 2: Prioritization of possible causes

Note: For this study “cause and effects” and “fault tree analysi”s was conducted for possible
causes with a severity level of 2 – 5.

Page | 14
5.4 Cause and Effect Diagram

Measurement Materials Method

Low
Discharge
Pressure of
EOG-Oilbird
Leak in the discharge FWP Pump
piping system

Impeller partially
clogged

Environment Manpower Machine

Page | 15
5.5 Fault Tree Analysis – Leak in the discharge piping system

Leak in the discharge

piping system

Holes in the
Column Pipes

External Corrosion

Wrong Material Galvanic Localized

Type Corrosion Corrosion

Figure 4: Fault Tree analysis for leak in the discharge piping system

Page | 16
5.5.1 Fault Tree Analysis – External Corrosion – Localized Corrosion

External Corrosion Occurs as a result of the

environment in which the Pump and
Column Pipes operate

Pitting and crevice corrosion formed

Localized Corrosion due to damage in the protective
coating

Poor Manual Handling Poor handling of Column Pipes

Techniques resulting in damaged coating

Damage of the
Resulting in exposed Column Pipes
Protective Coating
to the environment which is high in
chloride. The environment may also
setup a differential aeration cells and
pitting can initiate at the anodic site.
Lack of awareness
of the effects of
damaged
protective coating

Figure 5: Fault Tree analysis for Localized Corrosion

Page | 17
5.5.2 Fault Tree Analysis – External Corrosion – Galvanic Corrosion

External Corrosion Occurs as a result of the

environment in which the Pump and
Column Pipes operate

An electrochemical process in which

metals of different electrode
Galvanic Corrosion potentials corrode when in contact
with each other, in the presence of
an electrolyte (e.g. seawater).

Cathodic Protection Sacrificial Zinc anodes consumed

System failing to protect which therefore, allows galvanic
the Column Pipes corrosion of the less noble material
(LA Carbon Steel).

Sacrificial Zinc Anodes were changed

Lack of Preventive
Cathodic within the last 5yrs therefore,
Maintenance of the
Protection System preventive maintenance was done
Cathodic Protection
Inefficient on the system.
System

Figure 6: Fault Tree analysis for Galvanic Corrosion

Page | 18
5.5.3 Fault Tree Analysis – External Corrosion – Wrong Material Type

Occurs as a result of the

External Corrosion
environment in which the Pump and
Column Pipes operate

LA-C Steel and A 106 CS both contain

the same anodic index of 0.85 V.
Wrong Material Type Therefore, the material type will have
no impact on the rate of Galvanic
LA-C Steel instead of A
Corrosion (in the absence of Cathodic
106 CS
Protection) or the type and quantity of
Sacrificial Anode required.

Figure 7: Fault Tree analysis for wrong Material Type

Page | 19
5.6 Fault Tree Analysis – Impellers partially clogged

Impellers Partially Impellers partially clogged due to

Clogged Biofouling Biocide buildup.

Biofouling Bioside Biofouling Biocide buildup present

Buildup due to lack of Biocide Dosing

Lack of Biocide
Dosing

Figure 8: Fault Tree analysis for Impellers Partially Clogged

Page | 20
5.7 Summary of Findings

No. Possible Causes Root Causes

Lack of awareness of the effects of

1 Leak in the discharge piping system damaged protective coating
Cathodic Protection System Inefficient

2 Impellers partially clogged Lack of Biocide Dosing

Table 3: Summary of Findings

Page | 21
6 Action Plan – Recommendations
6.1 Action Plan – Localized Corrosion

Action Steps Responsibilities Timeline

No.
What Will Be Done? Who Will Do It? By When? (Day/Month)

Develop and implement a Standard

Operating Procedure (SOP) for handling
1 EOG 28th December, 2018
of pipe columns in order to minimize
damage to the protective coating
Train Maintenance Personnel on the
2 EOG 25th January, 2019
aforementioned procedure
Repair damaged column pipe sections
3 EOG/TOSL 14th February, 2019
and recoat all column pipe sections.

Table 4: Action Plan for Localized Corrosion

6.2 Action Plan – Galvanic Corrosion – Cathodic Protection System Inefficient

Action Steps Responsibilities Timeline

No.
What Will Be Done? Who Will Do It? By When? (Day/Month)

Review & Re-design the present

Cathodic Protection system (if Low
Alloy Carbon Steel is to be used as the
1 EOG 14th February, 2019
pipe section) to ensure a Mean Time to
Failure (MTTF) of 5 years (alignment
with NFPA Standard for FWP Pumps)

Table 5: Action Plan for inefficient Cathodic Protection System

Page | 22
6.3 Action Plan – Galvanic Corrosion – Wrong Material Type

Positive Material Inspection conducted by TOSL Engineering Ltd revealed that the flange sections
of the columns are 316 Stainless Steel however, the pipe sections presently in use are Low Alloy
Carbon (LA-C) Steel and not A 106 Carbon Steel as specified in the reference drawing – ACAD
FILE:833602-PUMP-GA.DWG.

According to “Engineering Edge, 2018”; Iron, wrought, gray or malleable, plain carbon and low
alloy steels all contain the same anodic index of 0.85 V (See appendix 8 for further details).
Therefore, the material variation between the reference drawings and “as found” have no impact
on the rate of Galvanic Corrosion (in the absence of Cathodic Protection) or the type and quantity
of Sacrificial Anode required.

Positive Material Inspection of EOG – Toucan FWP revealed that both the flange and pipe sections
of the column pipes are 316 Stainless Steel (see Appendix 9 for further details). The use of 316
Stainless Steel throughout the column pipes eliminate the possibility of galvanic corrosion due to
bimetallic coupling of different materials. It also eliminates the need for cathodic protection which
decreases the preventive maintenance time and cost and increases the Mean Time to Failure of the
FWP.

Action Steps Responsibilities Timeline

No.
What Will Be Done? Who Will Do It? By When? (Day/Month)
Change the material type of the pipe
1 section of the column pipe from LA-C EOG 25th January, 2019
Steel to stainless steel.

Table 6: Action Plan for wrong material type

Note: If recommendation 6.2 is being pursued then 6.3 shall be ignored and if recommendation 6.3 is
being pursued then 6.2 shall be ignored.

Page | 23
6.4 Action Plan – Impellers partially clogged – Lack of Biocide Dosing

Action Steps Responsibilities Timeline

No.
What Will Be Done? Who Will Do It? By When? (Day/Month)

Develop and implement a

biocide dosing program for the
1 EOG/TOSL 14th February, 2019
FWP column pipe and pump
sections.

Page | 24
Appendix 1 – ACAD FILE:833602-PUMP-GA.DWG

Page | 25
Appendix 2 – TOSL-PO4656/PMI-001

Page | 26
Page | 27
Appendix 3 – Build-up of biofouling barnacles on impeller

Build-up of
biofouling
barnacles

Page | 28
Appendix 4 – Minimum build-up of biofouling barnacles on the suction
strainer

Page | 29
Appendix 5 – Proper sealing of column pipe flanges

No signs of leakage at the

flange mating surface

Page | 30
Appendix 6 – External Corrosion leading to holes in two pipe column sections

Crevice Corrosion leading to cavity

1”x .75” in Column Pipe #8

Crevice Corrosion leading to cavity

2”x 5” in Column Pipe #4

Page | 31
Appendix 7 – Impeller #1 with excessive clearance at the bore area

Impeller #1 was found with

shim between shaft and bore.

Impeller Bore Dia – 1.940”

Shaft Dia – 1.937”.

Page | 32
Appendix 8 – Anodic Index of Metals
Anodic
Metallurgical Category
Index (V)
Gold, solid and plated, Gold-platinum alloy 0.00
Rhodium plated on silver-plated copper 0.05
Silver, solid or plated; monel metal. High nickel-copper alloys 0.15
Nickel, solid or plated, titanium an s alloys, Monel 0.30
Copper, solid or plated; low brasses or bronzes; silver soldier; German silvery high
0.35
copper-nickel alloys; nickel-chromium alloys
Brasses and Bronzes 0.40
High brasses and bronzes 0.45
18% chromium type corrosion-resistant steels 0.50
Chromium plated; tin plated; 12% chromium type corrosion-resistant steels 0.60
Tin-plate; tin-lead solder 0.65
Lead, solid or plated; High lead alloys 0.70
Aluminum wrought alloys of the 2xxx Series 0.75
Iron, wrought, gray or malleable, plain carbon and low alloy steels 0.85
Aluminum, wrought alloys other than the 2xxx Series, cast alloys of the silicon type 0.90
Aluminum, cast alloys other than silicon type, cadmium, plated and chromate 0.95
Hot-dip-zinc; galvanized steel 1.20
Zinc, wrought; zinc-base die-casting alloys; zinc plated 1.25
Magnesium & magnesium-based alloys, cast or wrought 1.75
Beryllium 1.85

Reference – “Engineering Edge, 2018”

Page | 33
Appendix 9 – TOSL-PO4656/18/PMI-008 Toucan FWP

Page | 34