Data Centers and the Like
The recent surge in data center and AI facility development across Pennsylvania is driven by the state’s “strategic” location, access to affordable energy, and a burgeoning tech workforce. However, this rapid expansion raises several concerns that could pose significant challenges for all Pennsylvanians.
Let’s explore data centers and their impacts throughout the Commonwealth. We will demystify the broad category of data centers and provide more context on their services, which have implications for the State. We will also explain why you should care about Pennsylvania’s aggressive approach to attracting this industry.
Data centers and AI facilities can be categorized based on their function, scale, and specialization. The general use of these facilities is online storage and creating digital infrastructure for businesses, websites, social media, and cloud storage. The need for these services is apparent, as we continue to move to more advanced technology and services, but the growth and demand might not be as immediate as companies want you to believe.
Here’s a clear breakdown of the different types of data centers:
Types of General-Purpose Data Centers
There are several types of Data Centers: Enterprise, Colocation Data Centers (Colos), Hyperscale Data Centers, Edge Data Centers, and Micro Data Centers. All of these facilities require high energy and resource inputs to run. Hyperscale facilities are a larger concern given their extreme size, supporting cloud computing and massive internet services. These facilities require up to a megawatt of power, almost 1% of Pennsylvania’s total energy load.
AI facilities
Often used synonymously are Artificial Intelligence or Supercomputing Centers (AI). These are purpose-built for large-scale AI training (e.g., large language models). Some facilities can also be Cloud Platforms, allowing users to train and deploy AI models on demand without owning physical infrastructure. Often part of existing cloud infrastructure but optimized for model training/inference. These facilities are usually more complex, with more infrastructure and require more energy and resources than the data centers listed above. Here are some additional resources to help clarify the scope of this field.
Impacts of Growth
Data centers, particularly those supporting AI operations, are highly energy-intensive. In Pennsylvania, the 71 existing data centers already account for about 3.2% of the state’s energy consumption. This demand is projected to increase significantly, with some regions expecting a 60% rise in peak load by 2030 due to data center expansion. Such increases could outpace the grid’s capacity, leading to potential reliability issues and higher consumer electricity costs.
Data centers have a substantial environmental footprint. Cooling systems require significant amounts of water, and the energy consumed often comes from fossil fuels, contributing to greenhouse gas emissions. The increased demand for water and energy for cooling may also strain local resources.
The financial burden of infrastructure upgrades necessary to support data centers is often passed on to taxpayers and utility customers. There has been a lot of conversation around charging “large load” customers by the Public Utility Commission, and the concern for stranded costs. These concerns include costs incurred by PA and investors abandoning projects after investment. This could disproportionately affect low-income communities, especially those located near data centers.
This rapid data center development often outpaces existing regulatory frameworks. Policies that ensure data centers are energy-efficient, utilize renewable energy sources, and do not unduly burden local communities are needed. Advocacy groups are calling for increased transparency and accountability in the planning and operation of these facilities. The Federal Energy Regulatory Commission (FERC) has already ruled that these facilities increase the cost of electricity for residential users so much that they must be charged more equitably.
Conclusion
While data centers and AI facilities could offer economic opportunities and technological advancements, their expansion in Pennsylvania must be managed carefully to mitigate potential negative impacts. Balancing growth with environmental sustainability, grid reliability, and social equity will ensure that these developments benefit the state without compromising its resources or residents’ well-being. The argument that there is an exponential demand for these centers is a tactic used to bullrush their development before Pennsylvania creates the structure necessary to protect its resources, communities, and our utilities.