Estimating Pipeline Probability of Failure Due to External Interference Damage Using Machine Learning Algorithms Trained on In-Line Inspection Data
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Abstract External interference damage is one of the main causes of pipeline failure reported in publicly available industry statistics from agencies such as the Canada Energy Regulator (CER) and the United States Pipeline and Hazardous Materials Safety Administration (PHMSA). Thus, failures due to external interference are often the most significant contributors to pipeline probability of failure in risk assessments and can play a significant role in operator decisions regarding risk-control expenditures, for example when it comes to the installation of additional impact protection, pipeline diversion or pressure restrictions. The probability of failure due to external interference damage can be estimated by combining the probability that damage occurs (i.e. that the pipeline is hit), the probability that the impact is sufficient to cause instant failure and the probability of degradation to failure, given that damage has occurred. Degradation to failure is assessed using industry standard engineering models (such as the limit state functions given in Annex O of CSA Z662-19 [1]). However, the key challenge is predicting where, when, and with what energy the external interference damage may happen. The prediction of a “hit rate,” or impact frequency, can often be subjective or based on statistics, which may not always be applicable or accurate for use on the pipeline under assessment. Top-of-line (TOL) deformation damage (dents) reported by in-line inspection (ILI) are a clear indicator of past external interference, which could have been introduced by third parties, contractors or the operator themselves. ILI data from ROSEN’s Integrity Data Warehouse (IDW) — which at the time of writing contains results from over 18,000 inspections — has been used to train machine learning models to estimate the frequency of external interference damage (per km-year). The distribution of dent sizes combined with pipe parameters is used to estimate a distribution of dent force. The following may all influence the likelihood and energy of external interference damage and may be considered as predictor variables in a machine learning model: • Local population density • Land use • Excavator types (typical bucket dimensions) • Frequency of crossings (road, rail, other services) • Pipeline burial depth • Additional impact protection • Pipeline markers and warning tape • Patrol and surveillance frequency • Operational control activities • Pipeline material properties This paper presents an approach to estimate the probability of failure due to external interface damage that use more accurate and justifiable impact frequency statistics, which are generated using worldwide ILI data and additional influencing factors based on pipeline exposure, resistance and mitigations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it