Back to all posts
Projects & Retrofits7 min read·October 25, 2026

What Should a Plant Automation Support Agreement Cover?

TB
ThynkBlox Automation Team
Automation Engineering

What should a plant automation support agreement cover?

A PLC and SCADA maintenance support agreement should define the equipment covered, when support is available, how a request becomes authorised work, and who owns each step towards restoration. Its checklist should separate response from resolution, explain remote and site attendance, and state exclusions and access prerequisites. It should also define who approves changes and restart, without promising uninterrupted operation.

Use the agreement to answer a practical question: when the plant calls with a fault, what exactly happens next?

Which systems and activities belong in the scope?

Build the scope around an agreed asset list. Identify the covered PLCs, operator panels, SCADA applications, computers, interfaces and relevant software versions. Name the production areas covered and record boundaries with electrical maintenance, mechanical maintenance, business systems and other equipment suppliers.

Separate fault support from planned maintenance. For planned work, specify the activity, frequency, deliverable and person responsible for arranging access. Avoid a broad promise to “maintain the automation system” without describing the work.

Use this PLC and SCADA maintenance support agreement checklist when reviewing the scope.

Agreement itemWhat to write down
Covered assetsEquipment, application and version identifiers
Included workDiagnosis, agreed maintenance tasks and restoration assistance
Service hoursCoverage window, time zone and holiday arrangements
Request routeApproved channel, required details and authorised callers
CommitmentsResponse, update and any qualified restoration targets
Remote accessConnection method, approval and session controls
Site attendanceDispatch authority, travel arrangements and site prerequisites
ResponsibilitiesIncident owner, technical owners and restart authority
ExclusionsWork requiring a separate instruction or quotation
ClosureAcceptance evidence, outstanding actions and updated records

For support following a retrofit, connect this list to the handover requirements in your functional design specification.

How does a support request become authorised work?

Name the people or roles allowed to request support and approve expenditure. Distinguish permission to investigate from permission to change a running system, order parts or send an engineer.

Define the route explicitly:

  1. Report: the caller provides the affected asset, symptoms, operating state and contact details.
  2. Accept: the provider acknowledges the request and confirms whether it falls within scope.
  3. Classify: the designated contacts agree its priority against written criteria.
  4. Authorise: the plant approves chargeable work, access and any proposed intervention.
  5. Assign: an incident coordinator records the next action, its owner and the next update.

An acknowledgement should not silently authorise unlimited work. Set any included allowance or spending limit, identify who can extend it, and explain what happens when the approver is unavailable.

Where immediate diagnosis is pre-authorised, define its boundary. It might permit reviewing logs while still requiring separate approval for a download, reboot or production interruption.

When is support available, and what does response mean?

State the service window in a named time zone. Explain whether calls outside that window are queued, handled through an agreed on-call arrangement, or subject to availability.

Define response as an observable action. An automated receipt, an engineer accepting responsibility and an engineer beginning diagnosis are different milestones. Say which one satisfies the commitment.

For each priority, record:

  • The operational condition that qualifies.
  • When the response clock starts.
  • Whether it runs outside service hours.
  • What pauses it and what evidence records that pause.
  • The update interval and escalation contact.

Keep severity separate from frustration. Define a production-stopping fault, a degraded process and a routine request in operational terms. Also specify who settles a disputed priority.

How should resolution commitments be written?

Separate a response commitment from a restoration target and a permanent resolution commitment. Define each term within the agreement.

For example, restoration might mean returning an agreed production function to service through an approved temporary measure. Permanent resolution might require replacing equipment, correcting software or completing further investigation.

If you include a restoration target, state the conditions and dependencies attached to it. Address unavailable parts, missing project files, restricted access and work owned by another supplier. Require progress updates and a named next action when a dependency blocks work.

Avoid treating an estimate as a guarantee. Ask the provider to record its assumptions and revise the estimate when evidence changes. The agreement should describe an accountable recovery process without promising continuous production or a fixed repair time for every fault.

What should remote support and site attendance include?

For remote work, specify the approved connection route, who enables access and who supervises actions affecting production. Define account ownership, session approval, activity records and how access ends.

Make the prerequisites explicit: a usable connection, appropriate permissions, accessible application files and a plant contact able to confirm equipment conditions. State what happens to service commitments while these are missing.

For site attendance, separate dispatch approval from arrival expectations. Record travel charges, expenses, minimum attendance charges where applicable, and any limits on engineer availability.

Site prerequisites should cover induction, permits, safe access, isolation arrangements and assistance from authorised plant personnel. Do not let permission to attend become implied permission to operate machinery.

Any proposed safety-related design change should be handled through a separate scope with a qualified assessor. Keep production restart authority with the plant's designated responsible person.

Who owns restoration when several teams are involved?

Assign one plant incident coordinator to organise decisions and communication. Then identify the owner of each technical action.

The automation provider might diagnose controller behaviour while plant maintenance checks field equipment and another supplier investigates an interface. The agreement should say who requests that assistance, who approves its cost and who tracks the handoff.

Require every handoff to include:

  • Evidence gathered and actions already taken.
  • The specific question the receiving team must answer.
  • A named owner and expected update.
  • Any operating restrictions still in force.

Define the condition for accepting restoration. Include the agreed functional checks, the person witnessing them and the person authorised to release the equipment for production. Closing a support ticket should follow that acceptance.

Which exclusions and change controls need to be explicit?

List exclusions in terms the plant can apply. Candidates include new features, capacity changes, unsupported assets, third-party subscriptions, replacement hardware and work outside the covered production areas.

Explain how excluded work proceeds. Require a scope, price or charging basis, authorisation and an assessment of production impact before it begins.

For software intervention, define backup capture, version identification, approval, verification and rollback arrangements. Agree how modified files, settings and credentials are returned to the plant.

Record temporary measures with an owner, operating limits and a review condition. A temporary bypass should never become an undocumented permanent change. Work requiring a shutdown should follow a separate plan; see planning a PLC retrofit around a shutdown window.

What would an authorised response look like?

Illustrative example: a line stops and the operator panel shows a communication fault.

The authorised maintenance lead opens a request and confirms the equipment state. The provider accepts the incident, agrees its priority and begins the diagnosis already included in the agreement.

The engineer proposes a restart of a covered computer. The plant's designated approver authorises the interruption after the relevant checks. If diagnosis instead points to a failed field device, plant maintenance takes ownership of that replacement while the automation engineer remains responsible for the agreed control-system checks.

The incident closes after the plant accepts restoration. Any unresolved cause remains a separate action with an owner. This sequence makes authorisation, technical responsibility and acceptance visible.

What common mistakes should you avoid?

Avoid buying “priority support” without a coverage window, response definition or escalation route. Avoid listing equipment without confirming that the provider has the files and access needed to support it.

Do not assume remote support includes site attendance, or that attendance includes parts. Keep emergency fault support separate from enhancement requests.

Also avoid using an ordinary maintenance agreement to cover a major redesign. Where the work changes operating sequences, interfaces or safety-related design, define a project scope and its acceptance requirements.

Frequently asked questions

Does an agreement guarantee uptime?

It should not be read that way. State the actual commitments and their conditions explicitly. Availability of support and availability of production equipment are separate matters.

Should parts be included?

Specify this directly. List any included stock, its location, ownership and replenishment responsibility. Otherwise, define who identifies, approves and purchases replacements.

Can the provider close an incident after a temporary recovery?

Define the closure rule in advance. Require plant acceptance, documented restrictions and a separately owned action for any remaining investigation or permanent repair.

What should happen when the agreement ends?

Require an orderly handover of current project files, change records, open actions and plant-owned access details. Assign responsibility for removing provider access.


Preparing a support scope? Review our services, then contact us with the assets to be covered, required service hours and the responsibilities you need to agree.

Next step

Want to apply this to your plant?

Every project starts with a fixed-fee assessment, and you can stop after the report.