Data Center Sustainable Rainwater Management

To begin, it’s essential to define sustainability accurately. It involves meeting our own needs without compromising the ability of future generations to meet theirs. These needs encompass natural, social, and economic resources. When talking about sustainability, we refer to the “Triple Bottom Line” or 3P’s, which are: When developing data centers, the journey toward sustainability begins with selecting the project site. However, the site selection process is influenced by various and multiple criteria, including but not limited to: In most of the countries, local codes and Environmental regulations exist to prevent development that disturbs or located near sensitive lands, which are but not limited to: Accordingly, during the pre-design phase, the data center sustainability engineer shall conduct site surveys and assessments of the available natural resources to develop an “Environmental Impact Assessment” report demonstrating that any development will not negatively impact any of the available natural resources on or in proximity to the selected site. One of the most important aspects and parameters to be considered during the early design stage is the site’s hydrological pattern and rainwater precipitation volume based on available historical data. And accordingly leveraging these patterns to reduce water consumption for cooling and humidification and to preserve the natural land cover hydrologic regime of watersheds. In this regard, the data center sustainability engineer shall adopt environmentally sustainable rainwater management systems such as Green Infrastructure GI and Low-impact Development LID strategies to retain, on-site, no less than 95^th percentile of rainwater runoff volume. Examples of these strategies include: Normally the data center sustainability engineer conducts a preliminary water budget analysis to support effective design decisions and potential integrative design opportunities, by gathering data to quantify the project’s potential non-potable supply sources such as captured rainwater, graywater from flow fixtures or condensate from cooling equipment. Following the preliminary water budget analysis, the data center sustainability engineer will conduct a complete life-cycle cost assessment for the different on-site non-potable water supply source alternatives considering the following for each alternative: The best engineering practices for water management start by eliminating the water demand whenever possible (Air-cooled chillers, for example) or by reducing it to the minimum by adopting low flowrate fixtures (e.g., WaterSense Label), selecting native plants with limited water irrigation requirements, and installing intelligent irrigation controls for landscape management. The following are the best practices for water management and/or conservation, starting from the top with the most preferable down to least:

Only 3% of the Earth’s water is fresh. Out of that 3%, slightly over two-thirds is trapped in glaciers. According to these figures, freshwater demand is expected to outstrip supply by 40% by the end of this decade. In the U.S., buildings account for 13.6% of potable water use, which is considered the third-largest category behind thermoelectric power and irrigation. Water stress and demand will continue increasing if governments, industry, and non-governmental organizations do not take strict and clear action. In addition, the energy required to treat water for drinking, transport it to and from a building, and later, treat it for disposal represents a significant amount of energy use not captured by a building’s utility meter. Thus, the water crisis is increasingly intertwined with global warming and the loss of biodiversity, with each reinforcing the other. A key driver of this phenomenon is global warming, which creates a global energy imbalance that intensifies the water cycle, adding approximately 7% more moisture for each 1°C rise in global mean temperature. Additionally, land use changes — such as deforestation, wetland depletion, land degradation, and infrastructure development — are now affecting precipitation patterns and the distribution of rain between green water (soil moisture/vapor) and blue water (runoff/liquid) flows. Given these facts and the mismanagement of water sources, it is imperative that all relevant stakeholders become involved to take corrective actions for serious and effective water resource management. Specifically, in the data center industry, designers and builders should consider constructing green buildings that use significantly less water than conventional constructions. This can be achieved by incorporating native landscapes that eliminate the need for irrigation, installing water-efficient fixtures, and reusing rainwater and wastewater for non-potable water needs. The Green Building Market Impact Report (2009) found that LEED projects were responsible for saving an aggregate of 1.2 trillion gallons (4.54 trillion liters) of water.

Data center facilities are energy-hungry and resource-intensive, consuming a huge amount of energy and water, mainly for cooling systems that maintain the required operating conditions. One of the main important sustainability metrics related to data center operations is Water Usage Effectiveness (WUE), which has become one of the hottest topics besides Power Usage Effectiveness (PUE). Significant progress has been made in the last 10 years to improve the Power Usage Effectiveness (PUE) and energy efficiency of DC cooling systems. Some of these efficiency gains have been achieved by using advanced energy-efficient systems. Yet, energy efficiency gains and improvements in PUE came at the expense of large water consumption, which contributes to stress on the water supply. Water Usage Effectiveness (WUE) is one of the main efficiency metrics used to measure a DC’s sustainability and efficiency, along with PUE and Carbon Usage Effectiveness (CUE). Thus, seeking alternative renewable water sources that serve the facility cooling infrastructure is currently considered one of the main objectives during the predesign stage. Furthermore, researchers at the University of California, Riverside, have found that between 5 and 50 ChatGPT requests can consume up to 500 milliliters of water. Google’s increased AI development in 2022 resulted in a 20% increase in water consumption compared to 2021 metrics; in the same period, Microsoft reported that it consumed 34% more water. By 2027, the amount of water AI will consume in one year worldwide is projected to be on par with what a small nation consumes. Consequently, the water consumption by AI data centers is expected to increase, and seeking alternative renewable water resources such as water harvesting is essential to reduce the environmental impact.  To date, only a few technology companies have taken steps to apply what might be the simplest, most proven and most promising strategy to mitigate water risks: catching rainwater from the sky. In this regard, rainwater harvesting and reuse is considered one of the most sustainable water sources serving the cooling systems, in addition to its economic benefits especially in the areas where municipal and water utility source is considered expensive. Some industry leaders are recognizing the potential of rainwater harvesting: According to research carried out by Savills, it is thought that a data center may use up to 26 million liters of water each year, on average, per megawatt of data center power, which might not be covered by harvested water. However, with climate change and rising costs of water, the benefits of a water harvesting system also increase. Data centers’ large, flat roofs are ideal for rainwater harvesting. A 50,000-square-foot roof can collect approximately 31,000 gallons of water from just one inch of rain. Many data centers feature roofs larger than 100,000 square feet, and some hyperscale data centers feature roofs up to a million square feet. Rainwater harvesting consists of the collection and storage of rainwater, usually from roofs and MEP yards. It also involves filtering processes to remove solids, organic materials, and sediments. This includes water treatment to neutralize nutrients and bacteria for evaporative cooling and/or humidification use. The level of treatment required depends on: