Essential Elements of Data Center Facility Operations

Overview

It sets out a risk-focused 'mission critical' operating philosophy and then twelve program elements: environmental health and safety, personnel, emergency preparedness and response, maintenance, change management, documentation, training, infrastructure management, quality, energy, financial management, and performance monitoring and review.

Practical detail includes shift coverage models: the highest emergency-response capability is at least two technicians per shift with electrical and mechanical expertise, 24x7, while lighter models use one technician on nights and weekends or less-than-24x7 with on-call. Minimum asset record fields are type, sub-type, description, make, model, size/rating, location ID, maintaining trade, serial number, install date, warranty expiry and planned replacement date.

A spares inventory should include components whose procurement lead time exceeds the maximum acceptable downtime of the associated system, re-evaluated annually. A MOP should include date/time, site/contacts, overview, predicted facility effects, supporting documents, safety requirements, risks/assumptions, step-by-step details, back-out procedures, approvals, sign-offs and feedback.

Training is organized into capability levels such as Basic, Intermediate and Advanced, with all tasks mapped to a level and annual recertification exams. KPIs are collected continuously, tabulated monthly and reviewed formally each quarter; a common pitfall is setting success and failure thresholds nearly identical. A table of common program mistakes closes the paper.

Key points

1. Twelve O&M elements from EHS and personnel through energy, finance and performance review.
2. Highest response capability: at least two technicians per shift, 24x7, with electrical and mechanical expertise.
3. Minimum asset record: type, sub-type, description, make, model, rating, location ID, trade, serial number, install date, warranty expiry, replacement date.
4. Spares: components whose lead time exceeds the maximum acceptable downtime, re-evaluated annually.
5. A MOP includes back-out procedures, approvals and sign-offs among its items.
6. Training uses capability levels with annual recertification exams.
7. KPIs are tabulated monthly and reviewed quarterly; near-identical target and failure thresholds are a common pitfall.

Using it in design

• Export every placed component as a CMMS asset row with the minimum fields, leaving serial number and dates blank for operators.
• Add a 'critical spares' flag driven by vendor lead time versus maximum tolerable downtime per system.
• Seed MOP and EOP templates from the single-line diagram, for example generator fails to start or chiller plant loss.
• In the KPI module, require separate target, acceptable and failure thresholds and warn when they are nearly equal.

Who should read it

Managers building or auditing a data center operations and maintenance program.

Related standards

• EN 50600-3-1
• NFPA 70E
• Uptime Institute Tier Standard: Operational Sustainability

Publication details

Authors

Robert Woolley; Patrick Donovan

Publisher

Schneider Electric

Year

2017

Document number

WP 196

Document language

English

Source type

Manufacturer

Access

Free to read

Licence

Copyright retained; citation only — the publisher’s terms apply to the original document

Citation