Managing complexity with intelligent platforms.
Our latest International Construction Costs Report 2024 identifies the construction assets of high-tech sectors, such as semiconductor production and data processing, as key markets with high growth pressure and even higher investment volumes. Efficiency in planning and execution can result in substantial added value. To illustrate this potential, we delve into the distribution of data center infrastructure.
Anyone who, like us, deals professionally with the identification of possible locations for new data centers, moves in a complicated, multifactorial decision-making space. A space in which not only the multitude and complexity of the criteria and framework conditions are impressive, but also their development dynamics over time. Some of these factors are expressed in "hard" physical measures and variables (e.g. safety distances and energy consumption), while others are based on empirical values from other market sectors and industries (e.g. public acceptance and authorization structures).
As is often the case in fast-growing, new markets with high demand, the location of data center infrastructure has so far developed in a rather agile and ad-hoc approach. The conventional process tended to involve due diligence checks and feasibility studies separately for each potential location, lacking any systematic pre-selection process. While this agile approach led to functional and acceptable facilities, it may not have maximized economic value and sustainability.
With total investments that can easily be in the high three-digit million range at one location, small disadvantages and frictional losses make a big difference here.
Hard and soft factors and criteria when searching for a data center location
These ‘wild’ early years are over. We are now seeing a growing demand for more systematic, structured and professional approaches, both on the customer front and in our global market analysis (... see also the high-tech and IT focus in our International Construction Cost Index 2024) What developers and operators of facilities have lacked until now, for example, is an integrated, intelligent platform that identifies potential locations in the form of digital white space mapping, and generates the best possible – and most promising – combinations for the respective requirement profile from the complex network of hard and soft factors and criteria, from "must haves / should haves / nice to haves". The platform, which is being custom-built for this purpose, is aimed precisely at this gap in the market. In order to understand the basic concept of the system, it is helpful to see a structured overview of the location factors for data centers (see illustration).
Location factors for data centers
(aggregated and generalized representation)
A closer, exemplary look at individual criteria clusters (without any claim to completeness) in their dynamics of change over time reveals the interesting degrees of difficulty and interdependencies of such an integrated approach:
• Data centers take several years to design, plan and obtain approval - these are long periods of time in the midst of a highly dynamic transformation phase for the entire energy supply infrastructure. It is therefore not only a question of reliably describing the status quo, but also of making reliable forecasts regarding the availability of energy and its price levels on site.
• The sustainability cluster shows similar development dynamics: developers and operators of data centers aspire to transition from being overlooked to leading the way in decarbonization efforts. Factors such as low sealing rates, heat coupling to local urban networks and the utilization of conversion areas are becoming increasingly important.
• There is also a lot of movement in the catastrophe risk category. The accumulation of extreme events caused by climate change makes site resilience a top criterion - especially in terms of the volume of investment at risk and the potential collateral damage to data for cloud providers' customers.
• Even the technical specifications of data centers are not as much a ‘fixed size’ as they appear to be at first glance. According to most forecasts, a significant proportion of the total data volume required for mobile applications will gradually be taken over by more compact edge data centers with their very own requirement profiles.
• If you take a closer look at the local market parameters, for example, rather poor forecasts for regional approval processes and public acceptance need not be a killer criterion. If all other parameters are right, it can make sense to invest in new, tried-and-tested methods of participation, stakeholder and communication management.
• Even if all the checkmark lists have been processed, local technical parameters can become a decisive hurdle - for example, due to geotechnically challenging subsoil. There is no substitute for interdisciplinary engineering expertise on site as early as possible.
Platform for structured location search and analysis for data centers
As we said - it's complicated. And the cooperation project is a promising approach to mastering this complexity. The nucleus of the platform is a synthesis of databases on energy supply systems at federal, state and local authority level with publicly accessible databases (e.g. on natural hazards), which are weighted according to their ‘level of confidence’ using a scoring system, and made available to users via a digital map-based GIS interface. In the current expansion stage of the system, the current and forecast supply of renewable energy on site, and the risks from natural disasters, are the most important key criteria - the goal of the platform’s development is the successive integration of all relevant location factors.
The exploratory discussions with customers show us that we are on the right track here. The platform is the organic and compelling continuation of our shared learning curves from the past. Our practical project experience with developers and operators has shown that the more interdisciplinary experience and expertise from fields such as decarbonization, conversion, resilience, infrastructure, energy efficiency, planning, approval management and project management are incorporated into the processes, the better the added value generated by location decisions in the data center sector. The platform has the potential to become the digital booster for this expertise.
It's complicated? Let's make it easy!
