You are on page 1of 7

Proceedings of the ASME 2022 14th International Pipeline Conference

IPC2022
September 26-30, 2022, Calgary, Alberta, Canada

IPC2022-87064

THE WHY AND HOW OF DATA INTEGRATION FOR INTEGRITY MANAGEMENT


PURSUANT TO API TR1178

Jackie Smith Cesar Espinoza Karen Collins


Principal of Integrity Principal Engineer, Head of Business Line,
Management Systems, ROSEN Education Systems &
ROSEN Houston, TX Services,
Houston, TX ROSEN Houston, TX

ABSTRACT United States – totaling over 5,500 mi of pipe segments and


Data integration is a foundational element and a regulatory ranging in diameter from 10” to 24” – were analyzed following
requirement for pipeline operators as part of an Integrity the core processes described in API TR 1178. The datasets –
Management (IM) program. Having taken effect in July 2020, ranging from hundreds to hundreds of thousands of integrity data
the Pipeline Hazardous Materials and Safety Administration’s records from different sources including a variety of in-line
(PHMSA) new Gas Mega Rule heavily emphasizes the inspection (ILI) tool technologies and service providers – were
requirement for the validation of data and records. As a result, interpreted to distinguish pipe features or references from
pipeline operators must face new challenges to ensure they are anomalies. Subsequent analyses were applied to the integrated
in continuous regulatory compliance. Data integration plays a datasets and delivered per the pipeline operator’s specifications.
major role in supporting pipeline operators and engineering
consultants as they adjust to the new requirements set forth in This effort greatly supported a seamless transition to
regulations. The Integrity Management (IM) regulations require achieve compliance with the new regulations and help to
pipeline operators to include the results of the latest as well as establish reporting requirements for technology and inspection
of previous integrity evaluations, including risk assessment service providers. The operator now has a complete historical
information based on integrated data from the entire pipeline. ILI database. As it is fully integrated into the GIS system of
Leveraging this data and the analysis thereof is required when record, the pipeline operator is able to transition into advanced
making decisions regarding the course of action to address analytics, helping them to spot trends and make sounder integrity
pipeline threats. management decisions long-term.

The American Petroleum Institute (API) Technical Report Keywords: ILI data alignment; data integration; corrosion
(TR) 1178, “Integrity Data Management and Integration,” growth assessment, Geographic Information Systems
provides a compendium of systematic methodologies,
recommendations and processes to spatially integrate and 1. INTRODUCTION
normalize integrity data to empower pipeline operators to In the United States, many pipeline operators are regulated
efficiently analyze integrity-related data to support their by the Pipeline and Hazardous Materials Safety Administration
integrity management decisions and programs. Through the (PHSMA). Recent updates to PHMSA 49 CFR (Code of Federal
execution of multiple post-ILI service projects, the significance Regulations) Part 192 Transportation of Natural and Other Gas
of this TR has been realized as a cornerstone to support pipeline by Pipeline: Minimum Federal Safety Standards make it
operators in their efforts to meet the requirements of new imperative to know details about pipelines that are regulated.
regulations and achieve compliance. These details equate to data. Data has become critical for gaining
knowledge about a pipeline operator’s assets. Data and
In the present case study, ILI results from various information derived from data have become critical assets for the
inspections gathered from over 179 post-ILI data integration pipeline operator and can be leveraged to improve safety and
projects for a mid-size pipeline operator located in the southern adhere to regulations.

1 © 2022 by ASME
Per API TR 1178 [1], data governance is the foundation for at a stationing or measure value than remember the associated
gaining insight from data, especially pipeline integrity data. GPS coordinate or ILI odometer reading.
Pipeline operators store and access data in many different ways. Furthermore, API 1178 brings awareness to integrating
For example PDF and Excel files may be placed on file share various datasets. Pipeline operators may provide pipeline
sites and not integrated with enterprise-wide systems, such as information such as High Consequence Area (HCA) segments,
GIS (Geographic Information Systems). Some pipeline operators class location designations as well as other datasets specific to
may keep data in silos where information is stored and accessed cathodic protection, above ground surveys, previous ILI and
only within their own departments as others may store data in excavation and repair sites which can be aligned and integrated.
enterprise-wide systems, such as a Geographical Information Datasets should be available in spreadsheet form or GIS format
System (GIS) or an Enterprise Resource Planning (ERP) system. to ease data alignments.
Pipeline operators often use industry approved data models such To achieve the required accuracy, the alignment process
as the Pipeline Open Data Standards (PODS™) or the Utility and must use a list of data points to align such as welds, valves, tees,
Pipeline Data Model (UPDM) from Esri™. taps, casings and sleeves, above ground markers (AGMs) and
It is highly recommended that pipeline operators, as bends. Pipeline operators have input on this list and can
competent stewards of data review API TR 1178 and use the comment on certain data, such as the quality of the AGM
recommendations therein as a foundation for data management. information or gaps, which as an example may direct the process
With the use of machine learning and predictive analytics to use more recent valve survey information first. In general, the
emerging in the pipeline industry, operators should continue to more points available for alignment the tighter the alignment.
improve the data management processes. Also, ILI to ILI alignments can be done, which are required in
Per API TR 1178, data integration is critical to successful corrosion growth assessments. ILI to ILI is a girth weld to girth
integrity management planning. Integration of new ILI with weld alignment and are simpler, it is more like comparing apples
previous ILI results and pipeline records offer more in-depth to apples whereas a GIS to ILI alignment is comparing apples to
understanding of pipeline integrity threats. Integrating and oranges. Also, since ILI to ILI alignment is a girth weld to girth
aligning multiple integrity datasets provides pipeline operators weld alignment, the amount of welds available to act as control
with more information and spatial awareness so the integrity of point allow for more precise alignments.
pipeline assets can be better understood, managed and analyzed. ILI to GIS alignments however are different. Typically, in
API TR 1178 addresses the full cycle of ILI and pipeline data GIS data pipeline records do not relate to specific girth welds.
management and integration activities for integrity purposes Other tie-points to align such as: Valves, Tees, Taps, Casings,
including: data quality, data integrity, data models and spatial AGMs and Bends must be used. ILI to GIS having less precise
alignment methods. It recognizes that, similar to the integrity of control points should be more thoroughly reviewed by the
the pipeline, integrity of data is critical for long-term use, analyst and should be managed by an established quality control
reliability, and key in making informed decisions. process.

2. Data Alignment 2.1 Quality Control


The first step for data integration is proper data alignment. Quality control is critical to ensure the ILI data is aligned
This will ensure that each ILI call, anomaly or reference type is with the GIS data. In order to ensure the quality of the alignment,
aligned or matched to its respective measure or engineering analysts developed a desired tolerance chart. This means that the
stationing on the pipeline centerline. API TR 1178 describes greater the distance between an alignment point of a matched
different approaches in the alignment for the purpose of pipeline GIS and ILI point, the less tolerance is given for that match.
integrity. Following API TR 1178 recommendations, pipeline Consequently, the more distance, the less tolerance the analyst
operators leverage linear referencing in relation to a pipeline would allow and would not match those two points.
centerline to store the pipeline asset information. Linear An example of what would fit in with the desired tolerance
referencing according to Esri™ is “A method for storing would include pipe bends called out from the ILI and pipe bends
geographic data by using a relative position along an already provided by the pipeline operator. In Figure 1 two pipe bends
existing linear feature; the ability to uniquely identify positions are set to match, the two on the left from the GIS are 49 feet apart
along a line without explicit X, Y coordinates. Location is given and the two ILI are 56 feet apart. This represents an 11 percent
in terms of a known linear feature and a position, or measure tolerance and is acceptable according to guidelines developed
along it.” [2] Those locations or measures for pipeline operators when doing alignments.
are stored in imperial or metric and are how pipeline features,
inspections, surveys and associated changes in attributes, such as
pressure, diameter, class location, pipe components, etc. are
managed along a digital pipeline. Points of interest along the
centerline, such as fittings, valves, launcher and receivers receive
a measure or station value, similar to how a house (a point of
interest) has an address on a road (think of the road as the
pipeline route). It is easier to refer to a valve or launcher receiver

2 © 2022 by ASME
execution for the final report. A structured data specification was
required to ensure proper exchange of crucial information and
the delivery of timely reports.
Data and information from the GIS was required to be
delivered with attributes such as the order of pipe segments,
beginning and ending engineering and cumulative stationing as
well as equations, class location designations, HCA information,
outer diameter, MAOP, percent SMYS, wall thickness, in service
date, seam weld types, coating information, product type,
method used to determine HCA and pipeline number. This data
FIGURE 1: DESIRED ALIGNMENT TOLERANCE and information was leveraged on a pipe joint basis for correct
data alignment as well as for inputs for the query calculations
There are occasions when alignment cannot be done due to required.
pipeline re-routing, replacements or removed features from the Features such as bends, flanges, markers, valves, launchers
GIS, as well as Out-of-date information from surveys and or receivers, taps, tees, and begin and end point of sleeves were
inaccurate or missing GPS points where the AGM information is required to be delivered with attributes of engineering
not complete. When these issues occur, updated records are stationing, bend degrees, descriptions and comments. An analyst
needed from the pipeline operator to make the alignments. segmented the pipeline data to create a record where each change
In addition to variance checking, the data alignment in attribute has a measure or point location along the pipeline
software includes quality assurance tools used throughout the centerline. The pipe properties used for segmentation included
alignment process. This includes a graphical alignment view and class location designation, date installed, MAOP, outside
two quality view graphs to ensure alignments are accurate. In diameter, wall thickness, and SMYS.
addition to the quality assurance tools, it is standard process for
another analyst to check the work and the overall quality of the 3.2 Query Development
alignment by ensuring any aligned points outside the desired The objective of the customized query development was to
variance are valid. support the pipeline operator in determining next inspection
dates or excavation dates for the metal loss anomalies and
3. OPERATOR CASE STUDY deformation defects reported. The pipeline operator also
ILI data interpretation is completed for alignments and required compliance assessments to be performed based on the
integrations in most post-ILI services for several small to large anomaly response criteria in 49 CFR 192 using standards such
sized pipeline operators. Analysts perform data alignments and as ASME B31G. The process developed assigns response
subsequent reporting for several pipeline operators who require timelines for anomalies detected by ILI run and provides the
that the ILI data be delivered with GIS pipeline references. As a pipeline operator with evaluations on metal loss and dent
case study, ROSEN collaborated with a pipeline operator with anomalies.
over 5,500 miles of natural gas pipelines in the southern United The queries utilized information from ILI results, feature
States. ROSEN provided these geospatial alignment services, box matching corrosion growth rate (CGR) and other
along with additional in-depth calculations and corrosion growth information uploaded into the software that runs the process.
analysis, to enhance the operator’s integrity management plans. The metal loss anomalies were grown over time using two
During the initial client engagement phase, it was clear the distinct growth rates. One was based on 75% of pipe life and the
pipeline operator had many challenges in their integrity planning other on CGRs determined through feature box matching.
processes due to datasets stored in isolation, including GIS data, In order to augment the process, certain ILI Information was
Excel sheets with ILI pipe tallies, alignment sheets and pipeline added to the base ILI run results to execute the queries. Extended
questionnaires. Additional issues included data inaccuracies, information includes the tool used during the inspection, box
duplication and redundancies. matching CGR results for depth per anomaly, box matching CGR
To help solve these challenging problems a program of results for length per anomaly, if available, dent strain analysis
customized queries executed on GIS aligned ILI results, results added as a percentage for the reported dent and the type
regardless of ILI service provider or tool technology was of the seam weld, required for execution of a customized query.
produced. Outcomes were integrated into a customized In collaboration with the pipeline operator’s integrity
spreadsheet deliverable that supports decision-making processes engineering personnel, the team collaborated to design and
related to ILI response, mitigation, and re-assessment intervals develop customized, parametrized queries based on the pipeline
according to the requirements outlined in the pipeline operator’s operator’s previous integrity process documents. As one may
Integrity Management Plan. expect, new requirements were discovered during workshops.
Integrity queries or calculations on the ILI and GIS integrated
3.1 Data Requirements datasets were designed with the intent to have partial results
The pipeline operator was responsible for proving pipeline available as an additional quality assurance measure. In total
data inputs necessary for integration to support the query fifty-five (55) customized, parametrized queries were created

3 © 2022 by ASME
and grouped into ten (10) categories. Where applicable, queries Pipe based on Parameters previously defined.
use requirements contained within pipeline standards from The Operator previously defined the
ASME and adhered to PHMSA regulatory requirements. The calculations in this section.
categories are included in Table 1.
Dent Each dent is assessed on a scenario basis
including dent location on the pipeline, such
Query Description as near a girth weld, metal loss and excessive
Category bending strain levels.

Parameter The Parameter queries provide base values Metal Loss of The o’clock position of the feature are
that are utilized in all subsequent tolerance Seam Weld compared to the seam-weld o’clock position
queries calculations to ensure the feature’s position is within the
prescribed proximity. The output scenarios
Tolerance The Tolerance Calculation is developed using also consider the seam weld segmentation
Calculation the tool tolerances provided by the Operator’s type.
specifications using the tables for 80%
certainty for anomalies located in the body of Next Based on the results of the assessments
the pipe or in girth weld of heat-affected zone. Inspection previously done combined with the
Date Remaining Life Queries a projection of the
Pressure The Pressure Calculation is based on the Next Inspection Date is provided.
Calculation pipeline operator’s specifications and
requirements for determining remaining
strength. These queries calculate the Safe and TABLE 1: OPERATOR DEFINED QUERY CATEGORIES
Fail (Burst) Pressures at the instant that the
ILI is reported making use of the Tolerance, Each query was evaluated to ensure it met the defined
Parameter and Intermediate queries. specifications and validated to guarantee the accuracy of the
Intermediate queries are the input factors for results. For validation purposes, a sample inline inspection
the more complex calculations which can be dataset was created with a small set of artificial features. These
features were manipulated into case scenarios to match the
viewed to ensure quality assurance
conditions defined in the query definitions outlined by the
Factor The Factor queries utilizes the results from operator. Validation procedures were performed for the queries.
the Pressure Calculation queries to determine For example, for the dent queries, dent features properties in the
sample dataset were manipulated to meet prescribed condition
the current Estimated Repair Factors and
while others were altered to fall right outside the threshold. In
Failure Pressure Ratio factors.
most cases the queries were validated based on the pipeline
Corrosion The Corrosion Rate calculation queries make operator’s remaining strength calculation specifications. Once
the validation results were consolidated, a semantic check was
Rate use of Parameter queries and segmentation
also performed by an integrity engineer who analyzed the final
values to estimate the Growth rates used later
deliverable report for accuracy and the correctness of the
on for the Pressure Calculation Queries (time algorithms.
based). The operator previously defined the Regarding the Next Inspection Date queries, it should be
calculations. noted that solely applying individual feature rates from box
matching is an approach that is often under-conservative. This
Pressure The Pressure Calculation queries (time risk should be considered when using results from the box
Calculation based) uses the Corrosion Rate calculation matched features to establish re-inspection interval or date. An
(Time based) queries to iterate over the years enabling the integrity engineer should consider this as part of the assessment.
calculation of the remaining life for each In fact, over the years, there have been some results where
defect. Each feature is grown yearly up to immediate excavations were flagged by the automated process
forty (40) years and is validated based on and those areas needed to be reassessed.
operator specifications compared to internal
software results. 3.3 CGA Requirements
To support the pipeline operator with their integrity
Remaining The Remaining Life queries will make use of decision-making regarding corrosion growth, a box-to-box
Life the Pressure Calculation Queries (time based) matching corrosion growth assessment (CGA) was also
to provide in years the remaining life of the performed. Following discussions with the pipeline operator, the

4 © 2022 by ASME
corrosion growth methodology of applying individually In addition, segmentation was defined based on the density
calculated corrosion growth rates was reviewed and assessed. of reported features along the length of the pipeline. For external
Recent publications have demonstrated that the application corrosion this process was semi-automated (where the
of individual CGRs can lead to inaccurate predictions of future automatically defined segments could be adjusted manually
integrity [3] [4]. An example of a subset of the results from the based on engineering judgement) using nearest neighbor
paper above are shown in Figure 2, which demonstrates the clustering (a modified DBSCAN algorithm [4].) For this method
paper’s findings of the application of individual corrosion a search window was defined for each feature, extending
growth rates was non conservative for a significant proportion of upstream and downstream by a specified distance suitable for the
the feature population compared to a segmented approach. pipeline being assessed. Each feature was considered moving
Therefore, a basic CGR segmentation methodology considered downstream along the pipeline. If the number of features within
appropriate to use with box matching in providing a the window exceeded a set threshold, the feature was considered
recommended CGR for each feature was included with the to be part of a “concentration”, along with all other features in
process. It is important to note that the results of each assessment the window. Each “concentration” ends when the feature density
were not verified by corrosion diagnosis or the comparison of falls below the threshold. Segmentation options were reviewed
signal data unless requested by the pipeline operator. for internal corrosion on a per project basis and typically, a single
segment was sufficient for the purpose of future integrity
assessment.
A characteristic CGR was estimated for each of the defined
segments, based on the historical depth changes of the features
therein. Depending on the number of features within each
segment along with engineering judgement, the characteristic
CGR may be based on a calculated upper bound (e.g. 95th
percentile) or the maximum measured depth change. Any
features considered to be statistical outliers were removed from
the assessment and considered separately.

4. THE PROCESS
The overall case study process developed for the operator
includes partial automation of the customer Integrity
Management Plan through software customization using data
management tools, python scripts, Excel macros and in-house
technical and engineering knowledge.
As shown below in Figure 3, the GIS data inputs and recent
ILI results were aligned by software automation. Further
software automation integrated the CGA results from the
engineering team. The final deliverables were created
automatically with some manual intervention before delivery to
the pipeline operator.

FIGURE 2: Comparison ILI Measured Depth vs Predicted Depth


when Considering Different CGR Estimation [1]

The estimation of CGRs is a critical input to any pipeline


integrity management program, and integral to determining
remaining life and in turn timescales for future repairs or re-
inspection intervals. The comparisons for the case study were
based on the most recent ILI data and one previous comparable
(equivalent inspection technology) ILI data set. The box
matching procedure and methodology consisted of a comparison
of the features reported in the two most recent inspections.
During this assessment, all reported corrosion was compared
using box matching software, which first performed alignment
of girth welds from multiple ILI runs. This process provided an
overview of how the reported feature depths have changed
between the two inspections. FIGURE 3: THE PROCESS FOR THE CASE STUDY

5 © 2022 by ASME
The deliverables to the operator included a standardized data in-depth decision making. Data can be integrated for displaying
integration report, inspection results, including CGA findings or each pipeline segment managed in the pipeline operator’s
box-to-box matching results and the recommended CGRs for Integrity Management Plan. Also, the data and information can
each corrosion feature, with a brief technical note describing the be integrated into the operators system of record and can be
methodology. accessed for PHMSA or state regulatory reports and audits
API TR 1178 discusses different types of data management quickly with traceability.
models beyond spreadsheets, such as object-relational data
modeling and dimensional modeling. The operator decided to
improve ILI delivery by requesting data in a standard data model.
The data model requested to implement is called The Utility and
Pipeline Data Model (UPDM) and is a template database
provided from Esri™ [5]. In this database, the ILI data can be
delivered for integrating back into a larger ILI database stored on
the operator’s servers. The UPDM database can serve as the
operator’s source of record for storing all ILI results over time
and can allow them to have a system of record. The operator can
in turn leverage historical ILI data for audits, trending and
performing analysis.
Lastly in the process, the pipeline operator requested that a
form be completed for record keeping purposes as part of the
requirements contained within their Integrity Management Plan.
The Excel form is populated by the analyst upon report FIGURE 4: DATA INTEGRATION PROJECTS
finalization before delivery to the pipeline operator.
Per API TR 1178, the pipeline operator recognized the
4.1 Process Improvements importance of integrating the GIS pipeline data and ILI data.
Over the years, ongoing improvements have been integrated Various checks could subsequently be performed after successful
into the above process. A year after working together, It became integration. For gas lines, for example PHMSA requirements
more efficient for the pipeline operator to provide the GIS in have been included for metal loss anomalies affecting the seam
shapefiles (a spatial file format) with the same data requirements, weld, at vintage pipe with low frequency ERW seam type pipe
instead of the within an Excel pipe properties and associated data on a HCA. The calculation for this query was done using a
listing. Asset information provided by the pipeline operator in combination of ILI and GIS data where the metal loss affecting
shapefile format included pipe properties, HCA segments, class the seam was derived from the ILI and HCA information, pipe
location designations, fittings such as such as tees and taps, vintage and seam type.
above ground markers (AGMs) and valves. In order to satisfy Further benefits of the data integration query were realized
the new requirement, an application specialist developed tools to when the seam type was unknown, and other conditions were
convert the shapefiles into formats for the internal software to met and a conservative approach was required. The process
continue the process efficiently. flagged the anomaly as an immediate repair which required the
Additionally, in 2020, AGMs were provided to improve the pipeline operator to take prompt action, such as make a repair.
accuracies of the inertial measurement unit (IMU) data on the Meetings were set up between the pipeline operator and subject
geometry tool inspections. Following guidance of the API TR matter experts involved with the engineering analysis where
1178, ILI runs that collect IMU data enable better accuracy when previous excavation, repair and cathodic protection information
mapping the centerlines of pipeline routes and provide valuable were reviewed for the pipeline operator to make decisions for
baseline data for the pipeline operator’s assets upon completion excavation planning and repairs.
of construction prior to pipeline operations. The benefit of using the integrated ILI results with the GIS
pipeline records integrated were that ILI anomalies that could
5. BENEFITS TO PIPELINE OPERATOR have potentially been missed from the repair plan were
Since the inception of the project in 2017, the pipeline identified. The process has successfully identified incorrect GIS
operator and ROSEN have collaborated on over 260 data records a few times which in turn led to the pipeline operator
integration projects with at least 150 of those including the confirming and correcting the GIS pipeline records.
corrosion growth analysis (CGA) results and roughly 22% of
those inspections being from other ILI service providers. Figure 6. CONCLUSION
4 shows the allocation of the projects per year separated by ILI In conclusion, API TR 1178 can offer in depth guidance to
service providers. Many benefits have resulted from this pipeline operators in implementing integrity management plans
collaboration including a streamlined and documented process, by providing detailed recommendations on data integration as
automated quality control procedures, and integration of CGAs noted. Data integration is the foundation for which other
and dent strain reports and support for dig sheet creation for more assessments, such as CGA’s and accurate responses to PHMSA

6 © 2022 by ASME
can be achieved. After five (5) years of collaboration, many
successful improvements have been implemented and the
requirements and deliverables will continue to be refined in the
years to come to keep up with changes in regulation,
technology and in the pipeline industry in general.

ACKNOWLEDGEMENTS
The concept behind the case study detailed in this paper was
based on an internal project at ROSEN between the operator and
the ROSEN Systems Implementation staff, conceived and led by
Karen Collins, Head of Business Line, Education Systems &
Services, with project support from Marcilio Torres, Senior
Application Specialist and Brett Jarvis, Data Integration
Specialist.

REFERENCES
[1] API Technical Report (TR) 1178, 1st Edition,
November 2017 - Integrity Data Management and Integration
[2]https://desktop.arcgis.com/en/arcmap/latest/manage-
data/linear-referencing/what-is-linear-referencing.htm
[3] IPC2016-64424: “Corrosion Growth and Remnant Life
Assessment – How to Pick the Right Approach for your
Pipeline”
[4] Ester, M; Kriegel, H-P; Sander, J and Xu, X (1996). A
Density-Based Algorithm for Discovering Clusters in Large
Spatial Databases with Noise.
[5] https://doc.arcgis.com/en/arcgis-
solutions/latest/reference/introduction-to-gas-and-pipeline-
referencing-utility-network-foundation.htm

7 © 2022 by ASME

You might also like