[Paper Review] Spatio-temporal load shifting for truly clean computing

Iegor Riepin, Tom Brown+1 more|arXiv (Cornell University)|Mar 26, 2024

Distributed and Parallel Computing Systems4 citations

TL;DR

This paper proposes a spatio-temporal load-shifting strategy using geographically distributed datacenters to achieve 24/7 Carbon-Free Energy (CFE) matching. By leveraging three signals—local renewable quality, wind generation correlation over distance, and solar radiation lags due to Earth's rotation—the model reduces 24/7 CFE costs by 1.29 ± 0.07 €/MWh per percentage point of flexible load, enhancing both cost-efficiency and clean energy utilization.

ABSTRACT

Companies with datacenters are procuring significant amounts of renewable energy to reduce their carbon footprint. There is increasing interest in achieving 24/7 Carbon-Free Energy (CFE) matching in electricity usage, aiming to eliminate all carbon footprints associated with electricity consumption on an hourly basis. However, the variability of renewable energy resources poses significant challenges for achieving this goal. We explore the impact of shifting computing jobs and associated power loads both in time and between datacenter locations. We develop an optimization model to simulate a network of geographically distributed datacenters managed by a company leveraging spatio-temporal load flexibility to achieve 24/7 CFE matching. We isolate three signals relevant for informed use of load flexiblity: varying average quality of renewable energy resources, low correlation between wind power generation over long distances due to different weather conditions, and lags in solar radiation peak due to Earth's rotation. We illustrate that the location of datacenters and the time of year affect which signal drives an effective load-shaping strategy. The energy procurement and load-shifting decisions based on informed use of these signals facilitate the resource-efficiency and cost-effectiveness of clean computing -- the costs of 24/7 CFE are reduced by 1.29$\pm$0.07 EUR/MWh for every additional percentage of flexible load. We provide practical guidelines on how companies with datacenters can leverage spatio-temporal load flexibility for truly clean computing. Our results and the open-source optimization model can also be useful for a broader variety of companies with flexible loads and an interest in eliminating their carbon footprint.

Motivation & Objective

• Address the challenge of achieving 24/7 Carbon-Free Energy (CFE) matching in datacenter electricity consumption amid renewable energy variability.
• Investigate how temporal and spatial load flexibility in datacenters can improve resource efficiency and reduce carbon footprint.
• Identify and quantify the impact of three key physical signals—local renewable quality, wind generation correlation, and solar radiation lags—on load-shifting effectiveness.
• Provide actionable guidelines for companies to co-optimize long-term energy procurement and short-term load-shifting strategies for clean computing.
• Demonstrate the marginal economic value of load flexibility in achieving truly clean computing at scale.

Proposed method

• Develop an open-source optimization model using PyPSA to simulate a network of geographically distributed datacenters with flexible workloads.
• Incorporate hourly electricity generation and pricing data for wind, solar, and grid mix across multiple European locations.
• Model temporal load shifting by rescheduling non-critical workloads and spatial load shifting by migrating workloads between datacenter locations.
• Use locational marginal carbon intensity and renewable capacity factors as key inputs to guide optimal load-shaping decisions.
• Apply a multi-stage optimization framework that minimizes 24/7 CFE matching costs while respecting datacenter operational constraints and grid interconnections.
• Validate results across diverse European locations and seasonal variations to generalize findings on load flexibility value.

Figure 1 : Illustration of spatio-temporal load-shifting optimisation problem faced by a datacenter operator.

Experimental results

Research questions

• RQ1 How does spatio-temporal load shifting reduce the cost of achieving 24/7 Carbon-Free Energy (CFE) matching in datacenters?
• RQ2 Which physical signals—local renewable quality, wind generation correlation, or solar radiation lags—most significantly influence the effectiveness of load-shifting strategies?
• RQ3 How does the marginal cost of 24/7 CFE matching change with increasing percentages of flexible load?
• RQ4 How do datacenter location and seasonal variation affect the dominance of different load-shifting signals?
• RQ5 What is the practical value of load flexibility for companies aiming to decarbonize their electricity consumption?

Key findings

• For every 1% increase in flexible load, the cost of 24/7 CFE matching decreases by 1.29 ± 0.07 €/MWh, demonstrating a clear marginal economic value of load flexibility.
• In Germany with partners in Denmark and Portugal, 24/7 CFE costs drop from 215 €/MWh (no flexibility) to 137 €/MWh with 40% flexible loads.
• For datacenters 380 km apart with similar wind conditions, 10% flexible loads yield 9.5% cost savings, increasing to 26.5% with 40% flexibility, peaking at 300–400 km due to wind correlation decay.
• Solar radiation lags due to Earth’s rotation enable effective load arbitrage across time zones, improving solar energy utilization and reducing CFE matching costs.
• The effectiveness of load-shifting strategies depends on datacenter location and season, with different signals dominating in different contexts—e.g., solar lags in summer, wind correlation in winter.
• Diminishing returns are observed with increasing load flexibility, indicating a need for strategic co-optimization of long-term procurement and short-term scheduling.

Figure 2 : Illustration of load flexibility concept: the absolute value of deviations between dispatched $\widetilde{d}_{t}$ [MW] and requested $d_{t}$ [MW] loads must fall within a certain flexibility range $f$ [%] for each hour $t$ .