Simply stated, the previously submitted noise study dated December 4, 2025 and reviewed by the Township’s acoustic consultant prior to the hearing illustrated compliance with all relevant ordinances and was on average 14 dBA lower than the daytime ordinance and 13 dBA lower than the nighttime ordinance requirements and at minimum (at the loudest receptor which is to the south adjacent to the PECO substation and on the opposite end of the property from Swedesford/Malvern Hunt) 9 dBA lower than the daytime ordinance and 5 dBA lower than the nighttime ordinance.

The sound studies we performed and had presented were conservative in that they: (i) presumed all impervious surface rather than grass and vegetation, (ii) assumed building IT/Server load at 100% capacity which is not a real-world operating condition given customer conservatism and unwillingness to max load to the redline. With a much more likely maximal building loading of ~75%, the chillers would run at lower capacity due to the redundancy of chiller count (~20% of chillers are redundant) and engineered safety margins in individual chiller sizing.

Given the design progression and better information regarding equipment selections and operating conditions, we have since re-run sound studies that update the total chiller quantity per building to 112 higher capacity chillers (reduced from 144 lower-capacity chillers) and also added ancillary equipment for completeness. We modeled noise at each receptor at the following operating conditions:

• Operation State 1: Most common operation (>99% of the time and for the life of the project, represented as 75% building utilization and peak summer conditions): 112 chillers operating at 55% with low sound operation.
• Operation State 2: Relatively uncommon operation (100% building utilization and peak summer conditions): 112 chillers operating at 79% with low sound mode operation.

We also added receptors near the Church Farm School, which indicate 50 dBA for Operation State 1, and 51 dBA for Operation State 2.

We will be submitting the resultant report to EWT for review by its consultant the week of February 2. On average all receptors indicate levels that are:

• Operation State 1 (most common): on average, 13 dBA lower than the daytime ordinance and 11 dBA lower than the nighttime ordinance
• Operation State 2 (relatively uncommon): on average, 12 dBA lower than the daytime ordinance and 9 dBA lower than the nighttime ordinance

Further, such results relative to the currently approved plan are on average ~2 dBA less overall.

Above and beyond Township requirements, we will commit to actively monitor noise at the property line for a minimum period of 3 years following construction.

No. The Township’s Noise Ordinance requires that our noise levels “be measured according to American National Standards Institute Methods for the Measurement of Sound Pressure Levels ANSI S1.13-1971 (R.1986).” That is what we have modeled. Noise studies are a standard engineering tool used by municipalities to evaluate expected noise impacts before construction when direct measurement is not yet possible. For this project, a licensed acoustical engineering firm used widely accepted software and methodology under conservative assumptions. The Township’s consultant confirmed both prior to and again at the hearing that the modeling employed by this consultant was consistent with industry best practice and that he would not have done anything differently.

Importantly, the study does not replace real-world compliance – regardless, the project must comply with applicable noise ordinances during construction and operation, with corrective action required if eventual measured levels exceed ordinance limits.

No. We have modeled noise at receptors at each of these locations and all levels indicate levels well below the ordinance’s requirements and near or below the levels modeled for the prior approved plan. Because sound levels decrease as they propagate over distance, locations farther from the site than the modeled receptors would experience lower sound levels than those shown in the analysis.

Yes. Soundwalls are still included and have been optimized for the new design.

In comparison to the previously approved Land Use plan, sound levels are ~2 dBA lower than the prior plan on average.

Also please note, as correctly indicated by the Township’s consultant during the hearing, the currently-approved plan used open cooling towers that employ large fans and a “waterfall” like mechanism whereby the noise-producing elements constantly run at 100% regardless of building loading all year round, unlike our air-cooled chillers that will modulate based on building loading and outside temperature. Said differently, with respect to noise, the noise modeled from our current plan reflects a worst-case moment on the hottest day of the year, while the noise modeled from the prior plan represented a constant condition every minute of every day year round. Moreover, these open cooling towers tend to generate a low, “whooshing / roaring” with splashing sound that carries far (low-frequency heavy) and are prone to imbalance and “wobble” creating noise effects more likely to be “felt” from a distance if resonating. The new plan uses air-cooled chillers with many fast condenser fans of composite propeller-type, which are lighter and less prone to imbalance and low frequency noise.

When ambient conditions are considered, it becomes apparent that the noise levels associated with our project are negligible in comparison to the noise levels that already exist. Noise levels are logarithmically additive: If the ambient condition at a location is 60 dBA and a 50 dBA noise source is added, the total noise remains at 60 dBA (i.e. the added source does not impact total noise level at all.) The fact that the Williams Pipeline pump station already produces considerable noise means that any noise produced by our project would be even less likely to be perceived. To put more precision around this, while we have not ourselves resampled ambient, the Township’s consultant previously measured and stated in his January 12, 2025 report that ambient noise levels at these locations are generally  “in the mid-50-dBA range” as compared to our most recent modeled levels that are generally in the low-50-dBA range.

Regardless of the extent to which certain residents feel the Township should update or modify its noise ordinance, the ordinance at the time of submission is the standard by which the Township is required to evaluate our amended application. And recall that our sound levels are materially below the ordinance anyway.

Noise studies evaluate compliance with the applicable ordinance, which regulates overall sound levels in decibels and does not include frequency-specific standards. Project review and compliance must be based on adopted, quantifiable, and enforceable requirements.

The project is legally bound to comply with applicable noise ordinances during construction and operation with corrective action required if eventual measured levels exceed ordinance limits. Sections 133-9 (A) and (B) of the East Whiteland Noise Ordinance set forth the requirements for enforcement of the Noise Ordinance as well as the penalties for violations of the ordinance.

No. We believe the movement of the office/loading complex to the north is better not worse with respect to the experience at Swedesford due to: (1) an ability to aesthetically enhance the northern façade in a manner consistent with a high-end office building with glazing and high-end materials as indicated in the renderings, and (2) moving such elements to the north enables a shift of the noise producing elements away from Swedesford. Note also that regardless of the buildings’ high end appearance, the buildings will be barely perceptible from any public right-of-way given vegetative screening as illustrated in the renderings provided.

The site/buildings will not be visible at all from the Church Farm School or the south/west vantage given that such location is over 2,000 feet away and there are trees/vegetation and adjacency to the PECO substation/easements.

The size and caliper (trunk diameter) of the trees we are purchasing pursuant to the enhanced screening/reforestation plan we have offered to implement are larger in every instance than those that are required by the ordinance and the existing in-place approved Land Use plan. For evergreen trees where 6-7’ heights are required, we are proposing a minimum of 8-10’ heights with certain elements as tall as 14’. For deciduous trees where a 1 ½” caliper (i.e. trunk diameter) is required, we are proposing a minimum of 3 ½” with certain elements as large as 5”. Please recall also that we have committed to installing substantially more trees and plant material than the currently approved plan requires.

No, not when basements are considered and the new building massing enables air cooling. The above ground portion of the buildings are bigger by ~250-300k sf per building, but when the fact that the prior buildings each had 250k sf sub-grade full-height basements is considered, the gross building sizes are nearly the same. The primary reason for the change in size and shape of the buildings and the elimination of the basements is that air-cooled chillers require more rooftop surface area. Evaporative chillers and related gear are more space efficient and can efficiently reside at grade but use enormous amounts of water (3M gallons+ per building per day) and are noisier. Air-cooled chillers use no water but take up much more physical surface area and must be placed in closer proximity to the load and their power sources – i.e. they require larger rooftop surface area than equivalent buildings that use evaporative cooling. The trade off for waterless cooling and quieter buildings is a different building geometry with larger above-ground area.

Please also note that the zoning allows for 1.86M square feet of data center on the site and we remain >200k sf below this legally permissible amount.

Generator emissions and diesel storage are regulated by DEP. This site resides in a region defined as a “non-attainment” zone by the DEP and thereby carries the most stringent emissions standard. As result, to the extent we install generators, we will be required by DEP to install “Tier IV” aftertreatment on each engine which uses a variety of methodologies including SCRs and associated DEF Systems which ensures minimal possible emissions, and there will be strict fuel consumption and/or run time limits. Also please note that generators are only all running in the event of a massive regional 230kV transmission outage on dual redundant circuits which has not happened in recorded history at the particular PECO substation serving our site other than for seven (7) hours during the Northeastern Blackout of 2003. Absent such an occurrence, generators are only running one or two at a time for testing on a monthly basis.

For this project, the emergency generator fuel will be stored in UL-listed, double-wall above-ground belly tanks (below each generator enclosure) designed and installed in full compliance with applicable NFPA, Fire Code, EPA, and DEP spill prevention requirements, including secondary containment, leak detection, and overfill protection. The system will also be subject to local fire marshal review and inspection. These belly tanks are exterior to the building and the building is staffed 24/7 with highly trained engineering personnel. These belly tanks are an industry standard for standby generators similar to what is used at hospitals and municipal facilities.

We have agreed to only run generators for testing during daytime hours and at maximum two units as was modeled for testing in the acoustic studies. Outside of this testing, the generators will only run during a regional blackout wherein the transmission grid and Planebrook Substation is completely down which has not happened in over 23 years (7 hours in 2003). Note that the overall run time for generators will be limited by the DEP and that this zone is in a non-attainment zone, of which has the most strict emissions standards.

We will not be discharging any glycol/biocides based on current building designs and deployment expectations, but such activities are regulated by agencies other than the Township.

This is a Superfund site that was cleaned up to standards for residential construction (the most stringent standards). All activities related to the Superfund status are governed and regulated by the EPA and not in the purview of the Township and we will, of course, fully comply and report to EPA as required. Further detail on the Foote Mineral Superfund site can be found at: https://cumulis.epa.gov/supercpad/cursites/csitinfo.cfm?id=0301103

No. Heat dynamics are not regulated by ordinance, however to ease concerns we have performed detailed computational fluid dynamic (CFD) modeling of the micro-climate issues relating to the air-cooled chillers on the roof. There is no notable change in heat conditions anywhere near or beyond the property line of our project due to any of the proposed buildings or systems we would install.

No. We are fully approved by EPA and all authorities to build a basement and could right now. We deliberately removed the basement to increase the above grade building area to make room for air-cooled chiller technology which consumes zero water to operate.

On the question of e-toxins, there are laws and regulations that will be complied with by our tenants who are typically large public corporations with extensive procedures and governance around how they remove, recycle and/or dispose of IT equipment.

No, there is NO blow down process for closed loop air-cooled chiller systems.  Our air-cooled chiller system uses a closed-loop water system, meaning the same water continuously circulates inside sealed piping. It is not exposed to the air, it does not evaporate, and it is not discharged to the environment. Before the system is placed into operation, it undergoes a thorough startup flush and cleaning process to remove normal construction debris such as dust or minor pipe residues. During this process, temporary strainers and filters capture any material inside the system, and that material is collected and properly disposed of. The water remains fully contained within the piping during cleaning and operation.  Once operational and stabilized, suspended solids in the system are typically maintained below approximately 5–10 parts per million, and the water is continuously filtered to remove particles much smaller than a human hair. The water is visually clear and periodically tested — typically monthly during the first year and quarterly thereafter — to confirm it remains clean and properly balanced.  A small amount of corrosion-inhibiting additive is included to protect the piping and equipment (similar to the treated hot-water heating systems used in public schools, hospitals, and commercial buildings).

Unlike cooling towers, a closed-loop system does not continuously add chemicals, does not produce mist or drift, and does not discharge treated water via a blowdown process. The water simply circulates within sealed piping and is maintained in a clean, stable condition for long-term equipment reliability.

Yes, all residential and commercial refrigeration equipment contains refrigerants, including the water-cooled chillers in the currently approved plan. The refrigerants proposed for use in the rooftop air-cooled chillers are not classified as hazardous materials. They are categorized under ASHRAE Standard 34 as A1 refrigerants, meaning they are non-flammable and of low toxicity. They also have zero ozone depletion potential (ODP) and comply with all current federal environmental regulations. These refrigerants are also used extensively in residential air conditioning systems, schools, hospitals, office buildings, and other commercial facilities throughout municipalities. They are materially different in hazard profile from substances used in certain industrial refrigeration applications (such as ammonia systems in large cold storage warehouses), which are toxic and subject to significantly more stringent hazard planning requirements.

Data center facilities are among the safest use-type with respect to life safety: we deploy pre-action smoke detection throughout the facility that detects particulates and provides early warnings of potential fire prior to the time that a typical smoke or heat detection device would alarm; we staff 24x7 with highly qualified facility engineers residing in an on-site network operation center (NOC) that are highly trained to respond within minutes to any potential alarm condition. We maintain and train on written and frequently reviewed disaster response procedures and business continuity plans and coordinate closely with the local fire department. This is required for our ISO 270001 certification. We use telemetry including building automation systems (“BAS”) and electrical power monitoring systems (“EPMS”) to real-time monitor thousands of points to catch mechanical or electrical issues before they cause failures and are typically bound by our customer agreements to maintain all gear on site to the highest possible standards. As such, insurance rates for data centers are typically significantly lower than for alternative building types and in our 20+ years of operation we have never had a data center in our portfolio cause any sort of safety issue to any member of any of the communities in which we have built. With respect to the gas pipeline, risks associated with the gas pipeline exist regardless of the development status of our project and are regulated and governed by relevant laws and regulatory bodies. The existence of data centers on our site, regardless of whether under the currently approved plan or the amended plan add no material risk (and we do not draw gas or connect to the pipeline in any way). A substantial gas transmission pipeline runs directly through dozens of data centers in “data center alley” and such condition is not uncommon in general given the co-occupancy of power and gas easements and, despite this, we are aware of no material related “disasters”.

There is no credible evidence that properly permitted data centers such as the one we are building harm people or children in any way. Period. 

Simply stated, these are buildings full of computers with standard commercial chillers on the roof that are utilized extensively throughout the County and emergency generators in sound-attenuated enclosures and screened equipment yards that almost never run. There is literally nothing in, on or around these buildings that could theoretically harm a person. There is no technology used here that isn’t used by dozens of commercial buildings and residences throughout the County.

These buildings in no way pose risks analogous to the risks associated with industrial manufacturing plants, chemical plants,  fracking operations, power plants, etc., all of which they are often falsely compared to.

These items have been often, and incorrectly, claimed to pose impact to public safety contrary to fact and any serious study:

• Generator Emissions: These are emergency backup generators that never run other than for minimal daytime testing (max two at a time) or a widescale grid outage that hasn’t happened at this grid location since 2003. EPA’s most stringent non-attainment zone standards apply to this site.
• Water Table Contamination: We are not using any process water and not discharging any water to sewer or storm other than that used for office-type functions such as bathrooms and sinks.
• Toxic Chemicals: Data centers do not manufacture or process chemicals. The only materials present are typical building systems such as batteries and maintenance/cleaning fluids (i.e. engine lubricant, cleaning liquids) common in almost all commercial buildings that are stored and handled by professional staff and regulated under programs overseen by the EPA and OSHA. Computer hardware is professionally decommissioned and properly disposed or recycled at end of life.

EMF/ Radio Waves/ Radiation: Decades of epidemiological research has found no credible evidence that EMF/Radio Waves/Radiation from electronics poses any human health risk. Regardless, EMF does not travel more than inches or feet from its source and does not propagate efficiently through air. EMF readings standing immediately adjacent to a data center wall are typically the same as background (i.e. imperceptible) and no different than standing next to an office building. Appliances like hair dryers and electric stoves produce far stronger EMF at close range and similarly are not dangerous in any way.

The currently approved plan employs open-loop cooling towers that would consume many millions of gallons per day of municipal water for cooling (roughly 3 million GPD per building).

Our amended plan deploys latest generation air-cooled chillers that do not require any water for cooling. Water will only be required at our project for typical office elements such as restrooms, sinks and water fountains, for typical fire sprinklers, and to initially fill the closed loop pipes within the facility during construction and perhaps with a de minimis “top off” on rare occasion.

No. Air cooled chillers can operate usually up to 131°F and above that temperature will shut off. This equipment does not have any physical connection for make-up water to evaporate. It is impossible.

The project will consume no well water and will not in any way reduce well water available in the region.

We are aware of no evidence that suggests that data centers of the nature we are proposing (e.g. architecturally appealing, significant setbacks, heavily screened, extensive noise mitigation, etc.) will have a negative impact on surrounding real estate values including residential homes.

There is in fact evidence that significant commercial ratables like data centers, which pay enormous taxes but require minimal municipal services relative to other uses like residential, result in lower real estate tax rates which positively affect real estate values, including residential.

Although there are many factors that affect home values including location, property condition, age, size, quality of school district, and macroeconomic conditions including interest rates, real estate taxes are widely considered the most significant community-specific economic factor and towns with strong commercial ratable bases often have the highest home values relative to nearby communities with higher tax rates. 

A 2025 study performed by the Schar School for Public Policy at George Mason University looked at the impact of data centers on for-sale housing values and concluded that: “The analysis fails to demonstrate statistical evidence that proximity to a data center negatively impacts housing values”. Please see study here: https://cra.gmu.edu/wp-content/uploads/2025/08/NoVa_DataCenters.pdf and abstract here: https://schar.gmu.edu/news/2025-11/study-home-prices-are-higher-when-house-near-data-center?utm_source=chatgpt.com

Toll Brothers, among the most experienced and prolific home builders in the country, just paid a record amount for residential property next door to our site with full knowledge of this entitlement for data center in immediate proximity.

We would also note that having an open and undeveloped environmental Superfund site next door is not helpful to home values. When finished, our project will be virtually invisible from public rights of way due to landscaping and reforesting and, once inside the new tree line, will look like a high-end office campus. We firmly believe that this development will not deteriorate property values and, as further context, have not seen a deterioration of property values in any neighborhood where we have developed a data center over the last 20+ years; in fact quite the contrary.

Issues related to power tariffs and market structures are governed by experts at the Public Utility Commission and other regulatory authorities.

The above notwithstanding, we understand that this is a sensitive and important issue and in a good-faith attempt to answer this question, offer the below. For those that read the below and really take the time to understand this issue deeply and seriously, we hope you appreciate that the soundbite that “data centers increase local power rates” is overly simplistic, sensationalist, and, as detailed below per a recent DOE-commissioned study, often empirically wrong.

Foundational facts/ relevant research relating to our project and power delivery/rates:

• This project will purchase transmission and distribution services from PECO pursuant to its “High Tension Tarriff”
• This project will purchase generation supply on the open market via direct power purchase agreements with one of multiple power producers that sell to large-load users in the PJM grid (for reference, the PJM grid extends from New England south to Virginia and west to parts of Ohio and Indiana).
• We are funding all improvements to PECO infrastructure required to provide power to our site via either cash payments or special provisions in the applicable tariff. Improvements are relatively minimal given our adjacency to the Planebrook Substation and we have funded a $28.3M letter of credit already which amount is subject to adjustment up or down depending on PECO’s final actual costs.
• Historically, this location in the grid used to support very significant power loads that has since been shuttered including power loads associated with the The Worthington Steel Rolling Mill, The 60+ buildings at the Foote Mineral processing plant on our property prior to the environmental cleanup, the Bishop Tube site on Malin Road, and The Johnson Mathey specialty chemicals company.
• PECO has provided a written statement stating that "PECO will provide electric service to your data center project via two 230 kV supplies coming from Planebrook Substation. PECO Service to your project will not adversely affect service provided to any other PECO customer."
• We understand the above statement is broad and could be interpreted to not specifically address power rates. Separately, PECO has recently completed an internal study on the impact of data centers on residential power rates within its territory. The results will be presented to each interested township over coming weeks/months. We would encourage the Township to reach out to PECO through their external affairs manager (Britnee Sanseverino) who can arrange to bring in PECO’s team to discuss their findings and any concerns they may have on Data Centers.
• The US Department of Energy recently commissioned Lawrence Berkeley National Laboratory to complete an empirical analysis of movements in power rates, including movements related to the addition of large loads like data centers, and concluded that, contrary to the cited concerns about large loads like data centers causing rates to increase: “…state-level load growth in recent years has tended to reduce average retail electricity prices… states with the highest load growth experienced reductions in real prices, whereas states with contracting loads generally saw prices rise.”
• Likewise, the following study published by EPRI, a highly respected international industry research organization, essentially came to the same conclusions: https://winwin.epri.com/
  •  “The analysis finds that states with faster load growth generally experienced smaller price increases, or even price declines, while states with flat or falling sales tended to see larger price increases”
  • “Under the right conditions, growing load can: (i) Lower average electricity prices by spreading sales-independent system costs over more kilowatt-hours; (ii) Improve the utilization of existing assets and enable more efficient grid planning; (iii) Accelerate deployment of clean electricity resources and emerging technologies; and (iv) Support better system operations and lower emissions.” (Note: “the right conditions” include addition of load in areas of the grid where pre-existing capacity exists and wherein upgrades are funded by the project – both of these conditions are satisfied by our project.)

More complex arguments relating to power rates and competition between utility territories:

• Transmission and distribution systems have enormous, fixed costs (corporate staff, maintenance fleet, office buildings, high-voltage wires, etc.) that do not change or change minimally as system load is added or subtracted. Mathematically, more kilowatt hours consumed within the territory serviced by these fixed costs means that that the fixed costs per kilowatt hour within such territory go down, not up, when more load is added. These “scale economies” affect all industries with high fixed costs and scale economies in power distribution are well understood by economists.
• Generation systems (i.e. power plants) as opposed to transmission and distribution systems are to a lesser degree governed by scale economy dynamics and are more governed by supply and demand of both capacity (quantities of plants) and fuel supply (coal, gas, nuclear, renewables, etc. etc.). In the Northeast, where most of the power plants that set marginal pricing are fueled by natural gas, swings in the commodity price of natural gas are by far the largest determinant of generation costs. It is also likely that demand from large loads like data centers place upward pressure on these generation costs until further generation capacity is added to the grid to meet this increased demand. But there is nuance here. PA has the 4th highest power generating capacity of any State in the US (behind TX, CA and FL). Virginia, by comparison, is 15th.  PA is currently exporting power to the PJM grid to service states like VA that have enormous data center demand in comparison to PA (VA data center operating capacity in megawatts is 20-30 times that of PA by most accounts). As a result, VA is getting the economic benefit of the scale economies in transmission and distribution referenced above (and the real estate taxes generated by its data centers as an aside) but net export states like PA (where relatively few data centers have been built to date) suffer equally from any upward shift in generation costs.
• Summarized simply, the above dynamics imply that, from a purely economic perspective, if data centers are going to be added one way or the other in nearby states within the PJM grid, a residential consumer benefits on a relative basis by having those data centers within their particular utility’s service territory so as to not pay all the grid-wide costs without getting any of the territory-specific benefits. The two states that added the most data center demand in the PJM grid from 2019 to 2024 were Virginia and Ohio where rates fell in real terms over this period while rates rose in states like PA, NJ, DE and MD that added almost no data center load during this period (see Figure 4 of the Paper referenced above).

There was no prescribed maximum power cap in the currently approved plan and no reason to believe the amended plan would result in more or less power consumption than the current plan given the fact that footprint is only loosely correlated to power consumption (i.e. power densities per square foot of deployment can vary wildly by application.)

No. Cybersecurity relates to malware issues that effect software and networks and have nothing to do with physical buildings and there is no relevance to EWT residents or businesses.

No.

To the extent Lithium batteries are deployed by our customers, their safe deployment and use is governed by the NFPA and the local Fire Marshall.

There is no meaningful connection and to the extent this building were ever decommissioned or demolished, such would be subject to all law and regulation at the time.

Yes. State law requires as much. In fact, Financial Security in the amount of 12.9 million dollars (as approved by the Township Engineer) has already been delivered to East Whiteland Township to secure the improvements associated with our approved development plan. To the extent necessary, additional financial security will be posted with the Township prior to recordation of our Amended Plan.