On October 31st at the Australia & New Zealand Awards Winners’ presentation, Collard Maxwell Architects [CMA] were announced as winners of the Sustainability Initiative of the Year category for their work on the delivery of data centers on Pacific Islands.
This Awards category was open to all forms of sustainability initiatives related to data center and cloud, including projects, technologies and solutions, operational processes, management, HR or strategic initiatives. The key to success lies in demonstrating best practices in industry sustainability over the preceding 12 months.
The Pacific Island data center project entered by Collard Maxwell Architects demonstrated a range of sustainability initiatives in part due to working in remote areas where there may be limited availability of key resources. The entry starts:-
“One common mistake with data center development in remote areas, such as in the Pacific Islands, is for a DC engineering team or future operator to propose copy-paste international design they are comfortable with, but that is oversized and not appropriate for the region or its available power. Another common mistake is within the business case, and is a failure to train local expertise in order maintain or replicate the design without heavy reliance on large international firms later on”.
The context of the project undertaken meant care was taken to design the outcome to meet the need:-
“The “Build it and They will come” approach is a common but key challenge for smaller communities. Building large white elephant facilities can take years to fill with racks in the context of a small Pacific island, and so could turn into a serious reputational issue for CEOs who may over-promise and risk under-delivering to their financiers and the public”.
This constraint extends to key design issues:-
“One key constraint to consider is, we know that most end-users are OK with 25ms latency to a primary DC located far away, so long as it is part of a larger ecosystem, like Sydney. So, with regards to any Pacific Island DC (unknowingly competing with Sydney), this leaves us catering to a smaller market of decentralized service providers with very specific needs; needs which impact on your infrastructure’s final design, size, staging, and delivery method”.
In order to execute the project, local expertise was trained through a “2N+1” team structure; and, the prototyping of new easily replicable and smaller DC hybrids were developed in a campus format, rather than as a single large, empty facility.
For this reason, CMA created a new simpler prototype for a cold shell with power and services, one which could then easily be wrapped with a design envelope by a local-architects, inspired by local vernacular. We believe that, because data centers in the Pacific Islands region must be maintained and upgraded also over time, that the right design delivery approach must be to reduce reliability on international consultants and equipment as much as possible.
Therefore CMA considers that a sustainable Pacific Island design does require a smaller easily replicable and affordable base buildings, with limitless architectural design possibilities provided by locals; and, ideally placed within a campus-style masterplan wherein some facilities might start off as a simple commercial tech headquarter, then slowly grow into a full-on small data center cluster, owned by locals, as/if the market grows. In the Collard Maxwell methodology, TWO architectural firms (one experienced and international, and one smaller and local), together with the addition of ONE team of peer-reviewer engineers, creates the ideal dynamic for a sustainable transfer of knowledge to other well established local firms.
The entry ends with some insight as to the role of the architect in Pacific Island data center design:-
“In designing a decentralized DC facility, or a business-park DC cluster, the Architect must lead the engineering team back to first principles. There is tremendous resistance to do so by the team, however, no design should ever move forward without a Lead Architect having detailed understanding of the CEO’s business case in the first instance. Seems logical, but it has been our observation that CEO’s are too often brushed to the side, turned off by complex engineering issues of the day. Don’t let it happen, this is a mistake”.