Industries · Energy & petrochemicals
The plant does not stop for a release window.
Refining, petrochemicals and upstream operations run continuously, and the cost of an unplanned stoppage is measured in production rather than in inconvenience. That single fact reorganises every technology decision underneath it — when change can happen, what a rollback has to be, and how much evidence a control needs.
This page describes how the sector works and which of Noble’s capabilities apply to it.
How the sector runs
Change is scheduled around turnarounds planned years in advance. A system upgrade that misses its window waits for the next one, which may be two years away, so sequencing is not a project-management preference — it is the constraint the whole programme is designed against.
The operational technology on the plant floor and the information technology in the office are separate estates with a boundary between them that is deliberate and defended. Anything that reads process data for reporting or prediction has to cross that boundary in a way the safety case allows, which is an architecture question long before it is a data question.
Regulators and auditors want evidence rather than assertion. A control described in a policy document and not enforced by a system is a finding, and the evidence trail is part of the deliverable rather than something assembled afterwards.
What the estate usually looks like
The shape most organisations in this sector are working with, and the parts of it that generate the technology work.
An ERP holding the operational record
Maintenance, materials, procurement and finance in one system, usually years into its life and carrying customisation that predates whoever is running it now. Most programmes in this sector begin as a question about that system.
Process data nobody can reconcile
Time-series data from the plant, production reporting built on top of it, and a finance system that disagrees with both. The reconciliation between them is where the reporting work actually is.
A hard boundary, and traffic across it
Segmentation between the control estate and the corporate one, remote access for vendors, and a growing list of things that need to read across it. Every one of those is an identity and segmentation decision.
Work that happens away from a desk
Inspections, permits, isolations and readings taken in the field, often where connectivity is intermittent. An application that assumes a network is an application that gets worked around on paper.
Which services this reaches
The four that a continuous-process constraint most often pulls in, and what each is answering.
- SAPUpgrades, conversions and rollouts sequenced against a turnaround calendar, with cutover and hypercare planned around a plant that keeps operating.
- Data & AnalyticsReconciling production, maintenance and finance to one definition, so a figure in a management report can be traced back to the transaction that created it.
- SecuritySegmentation between estates, privileged and vendor access, and evidence a regulator will accept rather than a policy statement.
- Managed IT ServicesSomebody operates it between turnarounds. Monitoring, patch cadence and a change process with a route back are the year-round work.
Technology typically specified
- Industrial computingEquipment rated for the environment it is installed in rather than for an office — temperature, dust, vibration and fanless operation.
- NetworkingSegmentation, site connectivity across a large footprint, and the wireless coverage field work depends on.
- Surveillance & accessPerimeter and area coverage, and controlled entry to areas where access itself is a safety control.
Start with the window.
The most useful first conversation in this sector is about when change can happen and what cannot stop while it does.