NIMBY - You're Not Putting a Data Center in My Back Yard!
- Apr 12
- 8 min read
Nerd Alert - Yes, it's going to be one of those posts - run away if you don't like technology blogs. But I'm going to make this one seem interesting even if you're a total technophobe.
You pick up your phone to look for a photo, turn on your computer to search the web, surf Netflix to pick a movie, or perhaps check your email. Does the Internet Fairy bring you this information? While there's always that possibility, your request is probably going to a database somewhere in a data center via some communications method and bringing it back to you lightning fast.
Were there always data centers? Nope. While the earliest data center was built in 1945 at the University of Pennsylvania, data centers didn't become popular until the 1960's when organizations like IBM dedicated facilities for its first supercomputers. These centers housed massive mainframe systems for military and intelligence purposes. They were housed in climate-controlled buildings. They were also super expensive and only governments, universities and large corporations could afford them.
In the 1970's and 1980's, large businesses maintained their database in dedicated on-premises computer rooms that required:
Heavy-duty power lines
Raised floors
Cooling systems
Dedicated Operators and technicians
These rooms were often in the basement of a building or campus and were noisy, often humid environments and looked more like this than the large data centers we see today:
And guess what? If the building or campus your business was in lost power, caught fire, had a backhoe take out your fiber connections or was attacked by a swarm of angry bees, you were SOL. (Technical term for not in good shape).
Introducing the dedicated data center. A dedicated complex that housed your very private, very important information in massive databases. Hospital needing a patient's chart information? It's here. A picture of your aunt Millie (the one with the wart on her nose)? It's safely stored here. That Netflix movie you promised your wife you'd watch with her but didn't really want to? Unfortunately, it's here as well.
There are several types of data centers:
Enterprise - owned and operated by a single organization to support their internal IT systems, giving them full control over infrastructure and security over the data.
Colocation - A shared space, power and cooling for multiple businesses, sharing and reducing the cost of the space while hosting your own servers and routers.
Hyperscale - Massive facilities (Think Amazon, Microsoft, Google) that are designed for scalable workloads, AI training and global cloud services.
Edge - These are smaller facilities close to the end users built to support real-time applications (Autonomous vehicles, IoT devices).
Micro/Modular - Compact, portable data center units that provide localized processing (often used in disaster recovery).
Confusing? Yeah, it's a lot. The health network I worked for had two small enterprise data centers that were close to our hospitals. Within the last ten years, they moved to two large colocation centers (I'd tell you where, but then I'd have to kill you...sort of thing) in the region.
What changed? Why the move from home grown computer rooms and enterprise data centers to large multi-tenant or hyperscale centers? The technology itself improved. The routers that move the data around got smarter. The carriers like PenTeleData, Verizon, LUMEN, AT&T, T-Mobile, among others got faster and became more reliable with self-healing fiber and load balancing networks that could redirect traffic to other centers in the world in real-time based on outages or clogs in the network.
Are there still outages? Fiber cuts? Software glitches? Absolutely. But the data world is getting better and with that better connectivity than ever before. It's a rare day that you can't pick up the telephone to make a call, turn on your computer and surf the web, or turn on the television to catch a show. When you understand that the AVERAGE household across the U.S. uses 20-40 GB of data daily, you understand the importance of a good network. Here's the breakdown:
Category | Share | Approx. GB/day |
Video streaming | 65–70% | 13–14 GB |
Social media | 10–12% | 2–2.5 GB |
Web browsing | 5–7% | 1–1.4 GB |
Cloud/app updates | 5–8% | 1–1.6 GB |
Gaming | 3–5% | 0.6–1 GB |
Audio streaming | 2–3% | 0.4–0.6 GB |
Video calls | 2–3% | 0.4–0.6 GB |
Mobile traffic now accounts for almost 64% of global internet traffic and short-form video watching from apps like Instagram, TikTok, Reels, X, etc. have caused an enormous uptick in data usage.
Let me talk about AI for a second. (Yes, this is me talking, not ChatGPT or whatever). Between AI and cloud computing, large language models, image generators and real-time analytics, the need for more computing power is greater than ever.
With this "need for speed" and huge database requests, the world needs bigger and better computing facilities to handle the load.
There are currently 39 planned or proposed data centers in the works for Pennsylvania this year. The largest clusters are planned mainly in Eastern PA, especially the Poconos, here in the Lehigh Valley, Lancaster, Carlisle and the Philadelphia metropolitan region. Some of the major ones are:
1. Amazon’s Pennsylvania AI Innovation Campuses
Locations:
Salem Township (near the Susquehanna nuclear plant)
Falls Township / Fairless Hills (Keystone Trade Center) Status: Announced / Planning Notes: Part of Amazon’s $20B PA investment.
2. Pennsylvania Digital I (PAX)
Location: Carlisle, PA Capacity: 1.35 GW total (300 MW by 2027) Developer: Pennsylvania Data Center Partners & PowerHouse Data Centers Status: Early development.
3. CoreWeave Lancaster County Data Center
Location: Lancaster County Capacity: 100 MW (scalable to 300 MW) Status: Early construction.
4. Project Gravity (Western Hospitality Partners)
Location: Archbald, Northeast PA (near Scranton) Size: 1.62 million sq ft Status: Planning.
5. Air Products’ Cetronia Road Data Center
Location: Upper Macungie Township, Lehigh County (Allentown region) Size: 2.6 million sq ft Status: Proposed.
Regional pattern
Based on the statewide mapping effort:
Northeast PA (Scranton / Poconos) — multiple new proposals
Lehigh Valley (Allentown / Upper Macungie) — major industrial‑scale campuses
Central PA (Carlisle, Lancaster County) — large hyperscale projects
Southeast PA (Philadelphia metro) — continued expansion around existing fiber hubs
Great, right? Eh, not everyone thinks so. There is a groundswell of opposition to these mega data centers being built in our area. Some of the complaints and concerns:
Environmental Impact - Data Centers consume a tremendous amount of electricity and water. Communities are concerned that the local energy grids and water supplies will be depleted or stressed leading to higher energy costs for our residents and reduced water availability for other uses. Remember, we're a big farming area and need this water.
The NIMBY (Not in my back yarders) - Construction of these massive data centers can last for years and generate, noise, dust, and heavy traffic. Additionally, the data centers themselves are often large, industrial looking buildings that take away from the charm and character of our small towns.
Economic and job concerns - Once built, a data center creates only a few permanent positions for the local economy for the amount of investment and land utilized.
Political/Ideological - People don't like tech giants in general - activists see data centers as a symbol of corporate overreach. Political parties like to demonize tech as prioritizing corporate interest over the public's welfare. Why are we cutting a tech company a corporate break on their construction when I can't get free (whatever) from the government to support my (whoever) instead?
Conversely, the advocates for bringing a new data center into a community say:
Job Creation - During that multi-year construction project, you're bringing construction and operations labor and jobs into an area, contributing to local employment and economic growth
Increased Tax Revenue - Data Centers contribute to increased tax revenues for local governments - those funds can be used to provide public services and infrastructure improvements.
Improvements to Infrastructure - Building a data center? You're going to need new roads, better power grids, enhanced telecommunication networks.
Economic Diversification - Support of local businesses and attracting other new business ventures.
What about the water?
One of the biggest concerns about a large data center is the amount of water it consumes. A large data center can consume up to FIVE MILLION gallons of water per day, which is the equivalent of the water use of a small town. Data Centers rely on water for cooling, which is necessary to keep things cool.
But guess what else uses a lot of daily water?
Agriculture (dominant global user)
Thermoelectric power generation (massive withdrawals for cooling)
Textiles & fashion
Chemical & petrochemical
Manufacturing, food & beverage, mining
In the scheme of things, a data center's water usage is minimal compared to what we're putting on our fields and crops. I'm not saying either is a good or bad thing. But we easily justify and understand that we need water for food. Let's face it, we're a hungry bunch.
There is a growing trend for data centers to REUSE their cooling water a few different ways:
Evaporative cooling uses some of the water by using whatever water that's left after cooling again and again - but because some of the recycled water literally disappears during the cooling phase you still need to replenish the evaporated water.
Internal closed-loop water systems reuse the water many times over and over but eventually, it does need to be replaced because minerals, bacteria and heat stress degrades the water.
Using reclaimed waste-water - about 120 data centers in the U.S. (including Google, Microsoft, Meta and Apple) work with utilities to take treated wastewater, clean it further and then use it for cooling. This saves over 530 million gallons of drinking water annually.
Zero-water cooling - this is the newest trend and where the industry is heading. Companies like Microsoft are launching zero-water-evaporation data center designs using chip-level liquid cooling and close-loop systems. The system uses water ONCE during construction and then never needs new water for cooling.
Less obvious is why we need big data centers. In my humble opinion, we've gotten used to the idea that the internet is that invisible entity that just works. Think I'm kidding? Turn off internet connectivity and within the first minute to several hours or days:
Communication collapse - No email, no messaging, no social media, video calls, cloud phone systems.
Financial transactions freeze - No credit card usage, ATMs, online banking, payroll systems.
Transportation snarls - Airlines, shipping, trucking logistics, GPS-dependent systems all fail
Emergency Services - No 9-1-1, dispatch centers, no hospital records
Business - Companies that rely on cloud services stop functioning. Supply chains break, factories that use internet-connected control systems stop production.
Government - National security impacted - intelligence, satellites, defense systems stop working
Stock Market - closes indefinitely.
Our internet service here at the house went down for a couple of days when someone hit a telephone pole on our street and we felt like we were Amish for a while. All I can say is, it's no wonder they have such large families.
If we don't continue to build data centers, how would we fare in the rest of the world?
If the U.S. stopped building new data centers, it would affect our global position, economy and our technological influence in the world in a very real way:
We would lose our leadership in artificial intelligence quickly - Countries that train, deploy and scale AI models dominate in defense, intelligence, biotech, pharmaceuticals, finance, manufacturing automation and global cloud service.
Our cloud economy would become stagnate - We'd lose billions in future tax revenue and U.S. cloud providers like AWS, Google and Microsoft would lose global market share.
National Security would weaken - threat detection, drone coordination, cyber-response would all slow down.
Scientific research would slow dramatically - Drug discoveries, climate modeling, genomics, fusion research, space exploration would all be affected.
Think of it this way. If we don't keep up with the advances afforded by new data centers, it would be the equivalent of refusing to build ports during a shipping boom.
If you haven't fallen asleep or moved on to watching Cat videos on TikTok by now, I hope you gleaned something from this post. As I see it, we really don't have a choice - we're going to need big new data centers to keep up with the world's demand. Whether it's in your back yard or someone else's, it's going to happen. But there's a bright future in the design, construction and deployment of how we do it.
Very informative. I had no idea!