1660317573518 Heroapm

The business case for advanced analytics and asset-performance management

March 29, 2021
An under-appreciated outcome of advanced analytics is a streamlined identification of data signatures.

By Alvaro Rozo, Uptake VP of application delivery 

The oil-and-gas industry is sitting on vast pools of under-tapped data.

A single well, for example, produces more than 10 terabytes (TB) of data per day. Every 400 km of pipeline inspected generates 1 TB. And with enhanced instrumentation because of smart sensors and automation controls, the amount of data streaming from critical assets is only accelerating.

The volume and velocity of data can be daunting. Much of it remains untouched after initial collection. By some estimates, just about 5% of all that data is used. Even less of that data is consistent, accurate, contextualized, or usable for different purposes. Only about 3% of companies met basic data quality and readiness standards, which severely restricts its ability to improve operations with advanced analytics—like the reduction and mitigation of unplanned downtime. 

For plant data to be as valuable as possible, the industry needs advanced analytics that distribute data access to the right people at the right time.

Advanced & predictive analytics

Advanced analytics tailor datasets for specific purposes to support decision-making, empower decision-makers to better manage machine performance and maintenance, improve reliability, lower costs, and reach business goals.

Click the graphic to join the webinar live or on demand.

Advanced analytics bring together data that documents current and historical equipment conditions and performance. That combination of OT data with context enables maintenance and reliability practitioners to identify dependencies and make learned assumptions.

Machine-learning (ML) algorithms work in this fashion and at scale—like an expert trained to aggregate data across sources and look for potential failure signals and usage patterns that lead to failures. The ML algorithm, designed or supervised by data scientists and equipment experts to detect those failures, uses that data to identify failure patterns on critical equipment. Based on that learning, the model can be a basis for predictive analytics.  

The preventative maintenance in place

Most organizations have set preventative-maintenance strategies on time- or use-based intervals. While routine, they do not offer much visibility into financial outcomes from maintenance.

Take unplanned downtime as a particularly costly example. Just 3.65 days of unplanned downtime can cost oil and gas organizations more than $5 million annually. It can be difficult to know without economic/performance benchmarks whether the plant is over-maintaining, under-maintaining, or maintaining equipment at just the right amount.

Decision-makers with an eye toward bottom-line results, sustainability goals, and regulatory compliance, however, can still leverage their OT data through advanced analytics to know how productive their plants are and how much more productive they could be. When presented at the enterprise level, such advanced analytics can become significant support for improved decision-making—especially with regard to unplanned downtime.

Distributed APM

An under-appreciated outcome of advanced analytics is a streamlined identification of data signatures that indicate pending equipment failure by those who know them best, not only from an asset-performance management (APM) perspective, but from an operating standpoint as well.

Historically, enterprise access to plant-level analytics, let alone advanced or predictive, was a pipe dream. In 2015, McKinsey found in a survey of oil and gas companies that less than 1% of the information gathered from rigs around the world was eventually distributed to decision-makers in the industry. Due to pre-modeling or networking constraints, data ingested from multiple OT systems at the plant often loses the context that makes it a valuable view into equipment conditions. Even before ingestion, many companies deal with data sprawl across those plant systems—data kept in SCADA systems, on-premise historians, CMMS/EAM, ERP, and other databases, including in spreadsheets on individual work computers.

Evaluating unplanned downtime across the enterprise

Advanced analytics powered by cloud computing is changing that sprawled reality for many businesses, sharing access to the right information at the right time across the organization. Departments that often work apart from other business functions—despite a shared objective to uncover cost savings and productivity improvements from the same critical assets—are able to share a contextualized view into plant conditions. 

Advanced analytics help oil-and-gas companies recover the untapped power potential of a plant. By detecting anomalous performance of critical assets, companies can minimize downtime—at the plant level with prioritized, high-value repairs. Decision-makers also benefit from standardized measurement and best practices—with a view of critical asset performance.

As an internal metric of productivity, lowering unplanned downtime puts in place more cost-effective and reliable maintenance strategies. With visibility at the plant, region, and enterprise levels, oil-and-gas companies can manage with the advanced analytics required to stay competitive. Advanced analytics, powered by the cloud, are giving companies cost-effective ways to evaluate unplanned downtime across their organizations.