Achieving Dispatchability in Data Centers: Carbon and Cost-Aware Sizing of Energy Storage and Local Photovoltaic Generation

TL;DR

This paper presents a scenario-based optimization method for sizing photovoltaic and energy storage systems in data centers to minimize costs and carbon footprint, considering local conditions and grid emissions.

Contribution

It introduces a location-aware sizing framework for PV and ESS in data centers that accounts for life cycle emissions and grid dynamics, improving sustainability and efficiency.

Findings

Optimal sizing varies significantly by region due to local conditions.

Some regions show high potential for carbon footprint reduction.

The method effectively balances costs and environmental impact.

Abstract

Data centers are large electricity consumers due to the high consumption needs of servers and their cooling systems. Given the current crypto-currency and artificial intelligence trends, the data center electricity demand is bound to grow significantly. With the electricity sector being responsible for a large share of global greenhouse gas (GHG) emissions, it is important to lower the carbon footprint of data centers to meet GHG emissions targets set by international agreements. Moreover, uncontrolled integration of data centers in power distribution grids contributes to increasing the stochasticity of the power system demand, thus increasing the need for capacity reserves, which leads to economic and environmental inefficiencies in the power grid operation. This work provides a method to size a PhotoVoltaic (PV) system and an Energy Storage System (ESS) for an existing data center looking to reduce both its carbon footprint and demand stochasticity via dispatching. The proposed scenario-based optimization framework allows to size the ESS and the PV system to minimize the expected operational and capital costs, along with the carbon footprint of the data center complex. The life cycle assessment of the resources, as well as the dynamic carbon emissions of the upstream power distribution grid, are accounted for while computing the day-ahead planning of the data center aggregated demand and PV generation. Case studies in different Swiss cantons and regions of Germany emphasize the need for location-aware sizing processes since the obtained optimal solutions strongly depend on the local electricity carbon footprint, cost and on the local irradiance conditions. Some regions show potential in carbon footprint reduction, while other regions do not.