Deadline pressures on new data centers are increasing, and that means increased consequences. In 2016, the Uptime Institute reported that 70% of Tier Certifications fail in the first round.  New construction for data centers is also speeding up , putting pressure on contractors to produce at a competitive pace.

With so many disciplines and subcontractors at work in a new data center project, contractors must insist on efficiency or risk a cascade of delays. You must pay particular attention to electrical testing deadlines. Power distribution safety and compliance issues require that you carefully and rigorously stay on task. In addition, close attention to data collection and reporting is the key to safe, efficient testing.

People: Dedicated to real-time data collection

New data center construction clients generally require particular formats for their facility data. An electrical testing partner that’s proactive about data will be able to adapt to any owner’s unique requirements, while also meeting your needs. The testing company should be able to show:

• Capacity for dedicated, on-site management of data collection
• Flexibility that meets contractor and client data/software needs
• Ability to coordinate timely data synchronization and upload

Quality control and verification procedures need to be integrated with the testing schedule, so that synchronization and upload are a clear priority. The capacity to correlate and check data for daily synchronization, especially with large testing teams on a compressed schedule, requires dedicated processes. Electrical testing companies experienced with data center clients should have the capacity to share results directly through a client dashboard or other interface. 

An experienced electrical testing company will get ahead of reporting requirements with careful planning to ensure that the data capture process fits every deliverable. This will prevent subsequent rework, such as a massive project of translating data or updating naming conventions. In the case of multiple needs, such as proprietary DCIM alongside regulatory compliance reports, testers should be capturing data within a structure that facilitates meeting all foreseeable needs.

Tools: Empowered for synchronization and accessibility

When the testing data is accumulated at an accelerated pace, or on a compressed schedule, data collection must solve efficiency issues without compromising quality control.

Data entry error isn’t trivial. Human error generally shows up in tasks like data entry at a rate of about 1%.  But a 2005 study found that entering a 10-digit number into a calculator resulted in a 5% error rate. In a safety context, a 1995 study of pilots’ entries in an aircraft flight management system found a per-keystroke error rate of 10%. 

Accuracy is ensured by doing away with manual transfers of data. Electronic data collection removes unnecessary steps by logging results directly from the testing device. This avoids transcription errors at the point of capture, speeds correlation, and improves efficiency for technicians. With automatic data collection, technicians can work quickly and safely. In addition, data is codified and organized immediately.

When electronic capture is not possible, the use of laptops or handheld devices in the field helps avoid excess paperwork and transcription errors. Testing companies should be able to provide software tools as needed or adapt to software required by contractors, clients, or commissioning agents.

Planning: Results in successful operations

A 2016 study of data center outages cites UPS failure for 25% of unplanned data center outages, with generator failure at 6%.  That totals nearly a third of unplanned downtime events attributed to failures of key power distribution backup systems. Integrated systems commissioning best identifies root causes of downtime issues before delivery. Adherence to energizing schedules and timelines keeps final commissioning schedules on track. Testing data must be organized and accurate so that integrated commissioning can properly analyze complex failure scenarios.

The amount of data that a new data center construction commissioning generates is enormous. Attention to rigorous data management by an electrical testing partner should also deliver:

• Baseline data for trending, benchmarking, commissioning, and sustainability certification
• Complete as-built information for operations, maintenance, and safety teams going forward

Proper data management includes a plan for each deliverable. Changes will happen. Therefore, electrical testing needs to be flexible, and the testing company should plan accordingly so it doesn’t negatively impact the schedule — or the budget.

Scalability is another key to success. If more resources are needed, providers should have the technicians and other resources on hand to complete testing and reporting on schedule. An experienced electrical testing partner will be able to expand service to meet needs throughout the commissioning process. An organized, thorough tester will have the data ready to advise owners and operational teams on any preventive maintenance required to preserve the power distribution system to specifications.

Conclusion

Electrical testing for new data center construction can be done safely while meeting strict deadlines. A trusted testing partner with data center experience will be able to demonstrate in-house safety policies that meet or exceed client safety programs, as well as OSHA and NFPA requirements. In addition, best practices in data collection and reporting produce efficient results, enabling contractors to meet deadlines.

Third-party testers put objectivity first and work within rigorous guidelines set by certifying agencies like NETA. When vetting a partner for electrical testing of a new data center build, contractors should consider accreditation and safety training, but should also look for expert on-site data management, pre-planning, and experience with the tools needed to deliver results on schedule. That way, you can avoid drastic penalties and protect your profits.

Cited works

1. https://www.networkworld.com/article/3142679/data-center/why-your-new-data-center-doesnt-work-the-sorry-state-of-commissioning.html
2. https://www.pwc.com/us/en/industries/capital-projects-infrastructure/library/data-center-construction.html
3. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.525.8487&rep=rep1&type=pdf
4. Panko, 2013 “The Cognitive Science of Spreadsheet Errors: Why Thinking is Bad” https://pdfs.semanticscholar.org/3d70/7aa11d9f9c96509fa718d0e80b141d4f2f52.pdf (or http://ieeexplore.ieee.org/document/6480329/)
5. https://www.ponemon.org/blog/2016-cost-of-data-center-outages