Find Vendors, Products, and Services
THE UPTIME INSTITUTE describes itself as ‘the Standard bearer for Digital Infrastructure performance’. The organisation’s Tier Standard created more than 20 years ago, has been used in data centre design, construction and operations across the world and is the globally recognized standard for data centre reliability and overall performance. To gain classification, datacentre topology and operational sustainability are assessed as performance criteria. Final certification confirms that the infrastructure has no weak areas, and the approved datacentre has worldwide accountability for excellence. Achieving Tier certification can be a lengthy process. Therefore, understanding a datacentre’s ultimate goal from the outset is necessary to ensure the right infrastructure is in place to help make any future transitions or developments as straightforward as possible.
The Uptime Institute certifies four Tier classifications to identify the anticipated performance of different site infrastructure design topology. From a UPS perspective, Tier I can simply be described as one UPS and one critical power path to the load: we call this an N configuration. Tier II brings N+1 resilience into the system where N = the quantity of UPS required to support the load and “+1” = the value associated to redundancy. There is still only one critical power path to the load.
This configuration offers a robust power protection solution which is unlikely to fail. However, both Tier I and Tier II infrastructure require a full shut down for essential maintenance and so clients demanding a higher level of availability may look for a Tier III or Tier
IV certified facility.
Tier III datacentres are required to have a minimum of 2N (N+N) within their infrastructure. This means two independent critical power paths each with their own UPS, ensuring that each path is concurrently maintainable, with at least one path always being available to support the critical load. Tier IV is a significant improvement on all other tiers as it requires redundancy on both sides plus multiple critical power paths: we call this 2N+1 ((N+1) + (N+1)). This concept provides a fault tolerant configuration with resilience and redundancy on both critical power paths, resulting in a fully concurrent, maintainable system.
For datacentres, the transition between Tier II and Tier III requires the most significant investment in infrastructure. So often, datacentre clients are concerned about a common point of failure within the overall infrastructure. but the latest UPS technology has removed all single points of failure from within the system.
The area that needs to be addressed is the common point of coupling or common node which is the shared element between the UPS and the downstream infrastructure. It is not possible to escape this in an N+1 infrastructure. However, common nodes are highly unlikely to fail because they are very strong reinforced cable connections. Electrical switch components are far more likely to expire so the N+1 or Tier II configuration remains particularly robust. Therefore, we encourage our datacentre clients who are considering making the move between Tier II to Tier III to design their infrastructure very carefully. Moving up a Tier, means upgrading the UPS: cables, switchgear, generators, and can increase footprint and energy usage needs too.
The investment needed can be significant. However, if designed with flexibility and future growth in mind you should only need to do this once. At the end of the day, this should be a strategic and economic decision based on the need to attract premium clients who are looking for higher levels of security, resilience, and maintainability.
We work closely with Data Centre Design and Build consultants and have experience of working with the Uptime Institute to provide appropriate UPS solutions for various Tiers. Therefore, we understand the process and the importance of careful consideration and correct interpretation of the relevant documentation.
Particular attention needs to be paid in relation to terminology. Terms such as active, passive or alternative need to be addressed and clarified as they can be perceived in different ways depending on the scenario. For example, if you asked several people from the industry what their understanding of a passive path is, you are likely to get several different interpretations. This can impact system design and so communication with the relevant parties involved is important to ensure the design is correct according to the relevant Tier classification.
It is also necessary to keep up to date with the evolving changes to topologies in the Uptime Institute’s documentation. Just like electrical regulations, certification needs to be kept current.
Understanding the ambitions of a datacentre early on, means the right infrastructure can be implemented from the outset. By selecting a scalable true modular UPS, a critical power protection system can be rightsized, flexible, resilient and futureproofed for the short and long-term. Modules can easily be added on a pay-as-you-grow basis ensuring the system continues to match actual load requirements from day one, day two and for the future.
True modular UPS systems are also designed with distributed architecture and redundant self-isolating capability. This means that no single module makes decisions for the whole system, and if any part of the system has a fault it is automatically isolated while maintaining the load. Installing UPS frames which are capable of housing far more modules than initially required, means the datacentre has the flexibility to add additional UPS modules when needed. Investment in additional UPS modules can be planned and introduced in incremental steps to help manage cost.
In order to move from Tier III to Tier IV classification it can be as simple as adding additional modules. Take for example a 200kW 2N (N+N) modular UPS configuration, with a frame containing 4 x 50kW modules protecting each critical power path. The addition of just one more module in each frame will take the resilience level of the UPS system to 200kW 2N+1 ((N+1) + (N+1)), all without the need to purchase additional infrastructure.
In contrast, a standalone UPS system can be costly in terms of Capex and Opex, as any growth could mean doubling the size of the whole infrastructure.
A datacentre will normally apply for certification once the relevant infrastructure is in place. However, if the criteria are not properly met, approval will not be granted. Any remedial works required to resolve the issues identified can be both time-consuming and
expensive to rectify.
Interpretation will need to be revised and refined to match the required classification, and agreement reached that the configuration resolves any potential challenges. It is far more than a box ticking exercise. The process can be intense and involves a collaborative approach to ensure a consensus of understanding is reached. It is a necessarily detailed process.
Gaining the required Tier certification means datacentres can offer a resilient and fully certified solution to attract new revenue streams and develop their client base. Selecting a flexible, true modular UPS system, and working with a UPS manufacturer that has the strategic goals of the datacentre in mind, can help on the journey to achieving the required Tier certification.
Originally featured in DCS Magazine November 2021.