A preventive-maintenance contract for a secured public-sector UPS fleet is not simply fifteen copies of the same inspection. It is a controlled operating programme in which technical work, site access, security screening, service-level commitments and calendar constraints must all be ready at the same time. GDF Technologies used that principle when planning an anonymized federal defence engagement covering fifteen UPS units. The award established the fleet size, defined response obligations and required screened personnel, with scheduled maintenance concentrated in a specified month. Customer identity, site locations, prices and unpublished contract terms remain confidential.
The case evidence records the awarded scope and the planning method. It does not record a completed maintenance result, measured reliability improvement or avoided outage. Those outcomes are therefore not claimed. The useful result is the control model: access and scheduling were treated as deliverables that could block the technical work, not as administrative details to resolve after award.
Operating context
The fleet served a secured federal environment where a technician could not arrive solely on the strength of technical competence. Personnel screening, site permissions, escorts where applicable, approved windows and communication routes formed part of the work boundary. The contract also carried defined service-level obligations between scheduled visits. That combination changed the planning question from “How long does one preventive-maintenance visit take?” to “How many authorized unit-hours can be completed inside the approved operating window without weakening response coverage elsewhere?”
The fifteen units were managed as a fleet, but each asset still required its own identity, location, configuration, maintenance history and access conditions. A fleet count alone does not establish the number of electrical work fronts, battery strings, bypass arrangements or shutdown dependencies. The work plan therefore needed an asset register and a site-by-site authorization map before technician dates could be promised.
The general maintenance baseline came from the GDF UPS maintenance protocols: pre-visit coordination, safety establishment, visual and environmental inspection, electrical measurements, battery testing, thermal inspection, functional checks where authorized, event-log review, restoration and a written service record [1]. The secured-site controls came from the field case and the UPS maintenance MOP guidance [2][3].
The work package is released only when the asset, access, operating state, technician and reporting controls agree. This is a planning diagram, not a site switching procedure.
Observed business risk
The largest early risk was not a known equipment fault. It was schedule failure caused by an incomplete dependency. A cleared technician can still be unable to work if the unit list is wrong, the escort is unavailable, the change record is not approved or the customer cannot release the planned operating state. A perfectly prepared MOP can still fail if the named technician’s screening is not valid for that location. A service calendar can also consume every qualified person during the planned month, leaving no reserve for an SLA call.
These dependencies interact. If one visit slips, the next unit may lose its window. If a maintenance activity exposes the load to bypass or removes reserve, another unit on the same distribution path may need to stay untouched until the first is restored. If a replacement part or battery is discovered to be necessary during inspection, procurement lead time may exceed the fixed month. The control plan therefore separates work that can be completed under the awarded preventive scope from corrective work that requires a new authorization, part, outage or return visit.
| Planning risk | Evidence required before booking | Release condition |
|---|---|---|
| Personnel access | Screening status, site-specific permission, escort rules | Named technician is accepted for the named site and date |
| Operating window | Customer calendar, change approval, affected-load review | Unit state and allowed tests are written into the visit package |
| Fleet capacity | Unit count, travel, expected duration, qualified coverage | Schedule includes contingency and preserves SLA response capacity |
| Technical boundary | One-line, bypass arrangement, battery isolation, known alarms | MOP matches the installed system and identifies stop conditions |
| Corrective findings | Parts route, approval authority, return-visit process | Preventive work does not silently expand into unauthorized repair |
Engineering qualification and constraints
Before the calendar could be treated as credible, the asset register had to distinguish UPS model, rating, serial number where permitted, battery system, monitoring interface, bypass arrangement and critical load. Previous reports and open recommendations were associated with the correct asset. Where documentation was incomplete, the uncertainty was carried into the plan as a survey or verification task rather than converted into an assumption.
The maintenance method also had to state what would and would not be tested. Battery impedance, float voltage, temperature, visual condition and connection inspection can normally be planned without a full autonomy discharge. Transfer testing, maintenance-bypass operation and load-bank work need explicit authority and an exact system procedure. An annual preventive visit must not be described as proof of rated runtime unless a controlled capacity test has actually been completed under a defined load and end voltage [1].
The MOP safety boundary is equally important. Every source, battery string, static bypass, maintenance bypass and possible backfeed path must be mapped. The approved operating state, roles, hold points, stop-work conditions, restoration checks and communication authority belong in the site package [3]. Generic maintenance language cannot substitute for the as-built one-line or the installed manufacturer’s procedure.
Options considered and decision basis
Three scheduling patterns were available. A unit-by-unit calendar is simple, but it can hide common dependencies and waste travel. A site-block schedule groups units at one facility, but it may remove too much reserve if systems share a load path. A fleet campaign concentrates technicians and reporting in the fixed month, but it works only if screening, materials and customer approvals are closed early.
GDF used a campaign model governed by unit-level release gates. The fixed month remained the planning anchor, while each unit had to pass access, technical and customer readiness checks before its slot became firm. Contingency capacity was preserved for discovered conditions and SLA calls. The schedule did not assume that fifteen units meant fifteen identical durations.
The campaign model provides coordination efficiency, but only when each asset keeps its own release gate and the fleet retains response reserve.
| Option | Advantage | Limitation | Appropriate use |
|---|---|---|---|
| Unit-by-unit booking | Clear ownership for one asset | Repeats access and travel effort; weak fleet visibility | Isolated units with unrelated windows |
| Site-block booking | Efficient for co-located units | Shared bypass or load dependencies can constrain sequence | Several units at one site with proven separation |
| Controlled fleet campaign | Aligns screening, tools, reporting and fixed month | Requires early data closure and contingency resources | Multi-site or secured fleet under common SLA governance |
Scope of work and control records
The planning package assigned a unique visit record to each UPS. That record identified the asset, site, allowed operating state, assigned technician, access approval, work scope, expected evidence and customer contact route. The fleet schedule then rolled those records into one programme view. This kept aggregate progress visible without losing the technical boundary of an individual unit.
The expected preventive-maintenance record included visual condition, environment, event history, operating values, battery measurements, thermal observations, alarms, test results, restored state and recommendations. Findings were classified by urgency and by the authority required to act. An observation that could be corrected safely within scope stayed in the visit. A battery replacement, intrusive repair, firmware change or bypass operation outside the approved package became a controlled follow-on action.
For organizations building a comparable programme, GDF’s Canada-wide UPS maintenance service describes the difference between preventive visits, emergency response and defined maintenance contracts. The link is relevant at the programme boundary: each service element needs its own inclusion, response commitment and authorization rule.
Implementation sequence
The campaign plan follows a six-stage sequence.
- Freeze the asset list and identify documentation gaps.
- Confirm personnel screening, site permission and escort requirements for each location.
- Build the unit-specific technical package, including the allowed state, test boundary and stop conditions.
- Reserve technicians, travel, tools and contingency capacity against the fixed month and the SLA.
- Execute each visit under its approved record, restore the unit and obtain customer acknowledgement.
- Reconcile the fleet register, open findings, corrective quotations and next due dates.
No stage is closed by email intent alone. The evidence is a named record: accepted access confirmation, approved change or MOP, technician assignment, completed service report, or customer acceptance. This reduces ambiguity when schedules move or personnel change.
Every gate produces a record. A date remains provisional until the asset, access, operating state and assigned resources are all confirmed.
Verification and reporting
Verification occurs at two levels. The unit-level closeout confirms the UPS was left in the intended state, alarms were reviewed, temporary controls were removed and the service report identifies any limitation. The programme-level closeout confirms all planned units were addressed or explicitly deferred, screening records remain current, open findings have owners and the fleet calendar shows the next obligation.
The SLA is not validated by the preventive visit itself. Response performance must be measured from the contract’s own event records, timestamps and exclusions. Likewise, the maintenance campaign cannot claim improved availability without a defined before-and-after measurement period. The defensible statement is narrower: the planning model aligns access, technical readiness and resource capacity so that scheduled work can be executed under controlled conditions.
Results and limitations
The verified field evidence establishes that GDF was awarded the fifteen-unit federal defence maintenance contract, that the contract carried defined SLAs and a personnel screening requirement, and that maintenance windows were planned for a specific month [2]. It also establishes the resulting house rule: clearance lead time and the service calendar are first-class scope.
The evidence does not establish how many visits were completed, whether any defects were found, whether availability improved, or whether a response target was met. Those facts are deliberately absent. The case is therefore a planning case, not a performance testimonial. The customer remains anonymous and no site, account price, serial number, clearance detail or unpublished contract term is disclosed.
Practical lessons for secured fleets
- Treat access as a technical dependency with an owner, evidence and expiry date.
- Keep one asset record per UPS even when the commercial contract is fleet-wide.
- Do not promise a fixed month until qualified staffing and SLA reserve are modelled together.
- Separate preventive scope from corrective work and define the approval route before the first visit.
- Write the allowed operating state and prohibited tests into each site package.
- Close the campaign with both unit reports and a reconciled fleet register.
Records required before the first firm date
A comparable secured-fleet programme is ready to book only when the current asset register, site access matrix, screening status, customer operating windows, approved test boundary, unit-specific MOP requirements, technician plan, spare-parts route and reporting template are available. If one of those records is missing, the date should remain provisional. The correct control is to expose the missing dependency early, assign it and preserve enough schedule reserve to protect both the maintenance campaign and the response obligation.
Sources
- GDF Technologies internal knowledge library, UPS Maintenance Protocols and Field Checklists.
- GDF Technologies internal field evidence, GDF Field Case Library: Engineering and Quoting Patterns, anonymized federal defence fleet entry.
- GDF Technologies internal knowledge library, UPS Maintenance MOP Safety Boundaries and Go-No-Go Criteria.
- GDF Technologies internal knowledge library, Preventive Maintenance of UPS Systems.
Director, GDF Technologies
LinkedIn: https://www.linkedin.com/in/christian-b-a22619a1/
GDF Technologies services every UPS brand covered here, across Canada. See our preventive UPS maintenance plans or call (514) 252-8324 for on site support.