Pinellas_County_Sheriff_Data_Center_Consult_2025(1)_compressed.pdf

Data Center Consulting Report

Pinellas County Sheriff’s Office – Data Center

Gap Analysis & Environmental Progress Review (May 2025)

Representative: Russell Schiebel (rschiebel@set3.com

Purpose

The purpose of this report is to evaluate the progress made since the 2024 environmental analysis of the Pinellas County Sheriff’s Office Data Center. This report provides a detailed review of updated test results, field observations, and laboratory analysis to determine whether previous recommendations were implemented effectively, and to identify continuing or new risks. Despite initial mitigation steps, the data center environment has declined, with silver corrosion escalating from ISA G2 (Moderate) in 2024 to ISA G3 (Harsh) in 2025.

Testing Objectives

• Electron Microscopy (SEM/EDS): Review degradation and elemental contamination on surfaces.

• Spectroscopy: Determine element levels and possible sources of airborne compounds.

• Laser Airborne Particle Counts: Evaluate particulate loads in compliance with ISO 14644 standards.

• ISA Slide (Coupon) Testing: Quantify corrosion film thickness for copper and silver.

• TO-15 Air Test: Identify volatile organic compounds (VOCs) contributing to corrosion.

• Temperature & Humidity Tracking: Determine correlation between moisture events and corrosion rates.

Pinellas County Sheriff – Consult Report

Visual Inspection

Inspection confirmed the continued influence of nearby external sources:

• Clearwater International Airport (~1 mile east): Aircraft emissions introducing SO₂ and

NOx.

• Largo Environment Services (adjacent): Wastewater and cooking oil processing releasing

H₂S and SO₂.

• Regional Landfill (Bridgeway Acres): Documented H₂S emissions contributing to corrosion risk.

The data center’s outside air intake remains set at 10%. The HEGA unit installed was undersized, not industrial grade, and lacking potassium permanganate (KMnO₄) media. As a result, sulfur compounds are bypassing filtration and aggressively attacking silver surfaces.

The steam humidifier remains a secondary contributor, pushing city waterborne contaminants into the environment.

According to the NIH (National Library of Medicine), a distance of less than 2 miles from a water treatment plant can cross- contaminate airflow with toxins. The type of toxins released from water treatment plants are , which are stated in ISA (International Society of Automation) standard for data center corrosive air and ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning

Engineers) as common corrosive influences on computer equipment such as servers and circuit boards, causing misreads, downtime and server equipment failures.

Treatment Plant Pollutant-Specific Distance Gradient Graphs:

Am Econ Rev. Author manuscript; available in PMC

2016 :Apr 27. Published in final edited form as: Am

Econ Rev. 2015 Feb; 105(2): 678–709.

marginal effects and ninety-fifth percentile confidence intervals from 8 separate the nearest operating toxic plant. The unit of observation is the monitor-plant dentified using plant openings and closings. In the regression sample, each adient can therefore be interpreted as standard deviations from the mean value.

pointwise standard errors in the figure are calculated using the delta method.

The Effect of Toxic Plants on Ambient Hazardous Air

Pollution Notes: This figure plots regressions of a single form of ambient hazardous pollution on a quartic in distance to pair and all regressions include monitor-plant fixed effects so the distance gradient is i pollutant has been standardized to be mean 0 and standard deviation 1. The distance gr Standard errors for the regression are two-way clustered on plant and monitor, and the Below each pollutant specific graph is a histogram, representing the number of monitor s at various distance bins from the plants in the sample.

Observation Analysis

The cabinet environmental logs (September 16 – October 16, 2023) and outdoor weather data from KPIE (St. Petersburg–Clearwater International Airport) together show why mini-condensation events and moisture-driven corrosion are plausible despite hourly indoor sensors appearing safe.

Indoor observations:

• Cabinet 15 Core temperature: ~58–60 °F

• Indoor dew point: ~46–50 °F

• Relative humidity: ~61–63 %

• Margin: Cabinet temperatures remained ~7–9 °F above logged dew points in hourly averages, so condensation was not directly recorded.

Outdoor observations (KPIE):

• Outdoor dew point routinely ~72–77 °F (mid-September through mid-October)

• Outdoor air therefore contained ~25–30 °F higher moisture potential than indoor measurements

Comparison Table – Indoor vs. Outdoor

Parameter Indoor (Cabinet log)

Outdoor (KPIE weather)

Implication

Temperature

(°F)

~58–60 ~75–85 (DB typical) Indoor is tightly cooled

Dew Point (°F) ~46–50 ~72–77 Outdoor has 25–30 °F higher moisture potential

Relative

Humidity (%)

~61–63 ~70–90 Indoor looks stable, but outdoor pulses can overwhelm

Condensation margin

+7–9 °F (safe at hourly cadence)

If outdoor slug mixes in, local Td rises >60 °F

A cold surface at 55–58 °F can easily dip below transient Td

Correlation to mini-condensation:

• Outdoor air at dew points in the 70s °F can leak in via dampers, doors, or pressure excursions. When mixed even briefly, the local dew point in front of a cabinet could rise into the low 60s °F.

• CRAC units delivering 55–58 °F supply air can cool metal surfaces below that elevated dew point. Even short pulses (5–20 minutes) produce beading, leaving residues despite not appearing in hourly averages.

• SEM/EDS debris with Na, K, Cl, and sulfates, and ISA coupons with silver film thickness (1,824 Å, G3 Harsh) confirm that wet-film events are occurring and enabling sulfur/halide chemistry.

Inactive or Overburdened Media Bed Consideration:

• A HEGA unit that is undersized, carbon-only, and without KMnO₄ is further compromised if media becomes wet or moist. Media saturation by humidity or short condensation episodes inside the bed can:

o Block pores and reduce effective surface area.

o Promote microbial or chemical fouling.

o Cause premature desorption or bypass of H₂S/SO₂.

• When outdoor-to-indoor dew-point gradients are this high, free moisture can condense inside gas-phase units as well, especially if airflow stagnates or if the bed cools below incoming dew point.

• An inactive or overburdened media bed becomes not just ineffective but a secondary source of contamination, since adsorbed gases can be re-released.

Summary:

The dew-point differential between indoor (~46–50 °F) and outdoor (~72–77 °F) demonstrates that even a stable indoor log can be punctuated by short-term high-moisture events. These events explain both the corrosion evidence and why undersized, moisture-sensitive carbon-only units fail. A proper industrial HEGA with pelletized KMnO₄ media is required to manage both the high outdoor moisture potential and the specific sulfur chemistry observed.

Conclusion

The Beyond Guardian Air (F159E) is undersized and chemically unsuited for a corrosive, critical electronics environment:

1. Airflow deficiency: Even with two units, the space remains under 2 ACH versus the 6–10

ACH target.

2. EBCT deficiency: Thin pad design yields EBCT too short for sulfur removal.

3. Media deficiency: Carbon-only cannot arrest H₂S/SO₂; KMnO₄ media is required.

4. Placement deficiency: Free-standing units do not integrate with room airflow.

Therefore, even with multiple consumer units, the data center remains exposed to corrosive sulfur attack, as proven by ISA coupon silver film growth (1,824 Å, G3 Harsh). Only an industrial HEGA system with KMnO₄ media, 1,600–2,800 CFM airflow, engineered bed depth and EBCT, and integration into the recirculation path can adequately mitigate the corrosion risk.

electronics environment:

1. Airflow deficiency: Even with two units, the space remains under 2 ACH versus the 6–10

ACH target.

2. EBCT deficiency: Thin pad design yields EBCT too short for sulfur removal.

3. Media deficiency: Carbon-only cannot arrest H₂S/SO₂; KMnO₄ media is required.

4. Placement deficiency: Free-standing units do not integrate with room airflow.

industrial HEGA system with KMnO₄ media, 1,600–2,800 CFM airflow, engineered bed depth and EBCT, and integration into the recirculation path can adequately mitigate the corrosion risk.

Surface Samples – SEM & Spectrum Analysis

Independent accredited laboratory testing (SEM/EDS) confirmed debris contamination on equipment surfaces. Elemental findings include:

• Sample 1 (MDA 2): Oxygen, sulfur, chlorine, potassium, chromium, nickel【52†source

• Sample 2 (Cabinet 15): Oxygen, sodium, aluminum, sulfur, chlorine, copper, zinc【

52†source】.

• Sample 3 (Cabinet 16): Carbon, oxygen, sodium, copper, aluminum, phosphorus, chlorine, potassium, calcium, silver, zinc【52†source】.

What These Results Mean

The presence of sulfur, chlorine, and sodium compounds within debris samples is a direct indicator that corrosive air compounds are settling onto and interacting with IT equipment surfaces. These airborne contaminants are consistent with external sources already identified

(wastewater facility, landfill, and airport), but their deposition inside critical equipment reveals how the facility’s current air treatment approach is failing.

Why This Matters for Equipment Reliability

• Sulfur (H₂S, SO₂): Sulfur compounds react aggressively with silver and copper, creating sulfides that interfere with electrical conductivity. In PCB finishes and connectors, this results in higher resistance, intermittent failures, and premature component degradation.

• Chlorine (Cl⁻ ions): Chloride attack leads to pitting corrosion, especially on aluminum and stainless steel, undermining structural integrity and creating localized hotspots of failure.

• Sodium and Potassium: Hygroscopic salts that absorb moisture, creating electrolyte bridges between conductive traces. This greatly accelerates electrochemical corrosion on fine-pitch circuit boards.

• Nickel, Zinc, Chromium: Their presence within deposits suggests breakdown of plated components or internal structural parts, meaning the corrosion is progressing beyond surface contamination into actual material degradation

Localized vs. Migrated Contamination

While the SEM/EDS and coupon results strongly support airborne sulfur and chloride migration, inspection photos and debris distribution also suggest localized moisture events may be amplifying the effect:

• Localized moisture (condensation, leaks, humidifier carryover): Deposits can form when water evaporates, leaving behind salts (Na, K, Ca, Cl, S) that mirror airborne contamination signatures.

• Airborne migration: ISA coupon data showing disproportionate silver attack (Ag 1,824 Å

vs. Cu 447 Å) confirms systemic H₂S/SO₂ penetration into the environment.

• Most likely hybrid: The facility is experiencing both: airborne sulfur/chloride ingress driving baseline corrosion, with localized water films accelerating deposition in specific hotspots.

Supporting Science:

• ASHRAE TC 9.9 (2011) notes that “surface moisture, even micro-condensation, acts as the electrolyte bridge that allows corrosion reactions to proceed.”

• Mazurkiewicz (2006) documented that reduced sulfur gases plus high humidity drive creep corrosion, but that leaks and condensation can intensify localized attack beyond what air concentrations alone predict.

Correlation with Airborne Compounds

The SEM/EDS debris chemistry aligns directly with the ISA coupon and TO-15 results:

Surface finding

(SEM/EDS)

Likely airborne precursor(s)

(ISA/TO-15)

Reaction pathway

Equipment risk

Ag + S → Ag₂S (silver sulfide)

H₂S / SO₂ Sulfidation;

tarnish film grows under humidity

Contact resistance, connector intermittence, PCB surface finish failure

Cu + S/O → Cu₂S / CuSO₄

H₂S / SO₂ Sulfidation and sulfate formation

Creep corrosion, overheating, premature board failure

Cl⁻ with Na⁺/K⁺ Aerosols, humidifier carryover

Chloride pitting, hygroscopic electrolyte films

Pitting, leakage currents, connector fretting

Hygroscopic salts (Na, K, Ca)

Sulfur oxides + humidification byproducts

Absorb water, form electrolytic bridges

Leakage, soft errors, corrosion under bias

Ni/Cr/Zn particulates

Internal plating breakdown + corrosive gases

Secondary indicator of progressing attack

Connector/solder joint degradation

Scientific Backing:

• “Principal product of HS⁻ and silver is silver sulfide (Ag₂S)” .

• “Locations with an ISA Class G1 environment for copper but where the silver corrosion rates are 10–20× higher” .

• “Accelerated corrosion of printed circuit boards due to high levels of reduced sulfur gases” (Mazurkiewicz, 2006) .

• “Keeping the relative humidity below about 60% will keep the leakage current… in the acceptable sub-µA range” (ASHRAE TC 9.9) .

Why Immediate Mitigation Is Required

This evidence confirms that airborne corrosive compounds are actively migrating into and damaging live equipment surfaces. The chemistry seen on the hardware is the natural endpoint of the air chemistry measured by the coupon and TO-15 testing. Without replacing the current undersized filtration system with a proper HEGA unit containing KMnO₄, the facility will continue to see silver-dominated attack that will shorten MTBF across all major electronic systems.

The chemistry is consistent with external airborne contaminant migration, but localized moisture intrusion is compounding the severity in certain equipment zones. Both vectors must be addressed to stabilize the environment.

Air Particle Counts

Laser airborne particle counts were performed consistent with ISO 14644 standards at the 0.5 micron size range. Findings remain in the same range observed in the prior 2024 report:

• 0.5 micron particles: elevated, consistent with human traffic, dust, and minor external intrusion.

• 1.0 micron particles: within moderate levels, comparable to previous tests.

• 5.0 micron particles: low but still detectable, similar to 2024.

Interpretation:

• Particle counts remain moderately above optimal for a mission-critical data center but are not substantially different from prior findings.

• The results indicate that particulate control measures (HEPA) are functional, and particulate levels are relatively stable year-over-year.

• The environmental issue driving equipment corrosion remains predominantly gaseous

(sulfur and chlorides), not particulate.

ISA Air Corrosive Levels – Coupon Testing

2024 Results (6/17/24 – 7/17/24, Coupon #182898):

• Copper: 839 Å (ISA G2 – Moderate)

• Silver: 421 Å (ISA G2 – Moderate)

2025 Results (5/9/25 – 6/9/25, Coupon #195141):

• Copper: 447 Å (ISA G2 – Moderate)

• Silver: 1,824 Å (ISA G3 – Harsh)

Discussion

Room volume is ~1,400 ft² × 10–12 ft ceiling = 14,000–16,800 ft³. For corrosive gas control, industry guidance calls for 6–10 ACH recirculated through the sorbent bed. This equates to a required airflow of ~1,400–2,800 CFM.

The Aerus Beyond Guardian Air (F159E) delivers only 75–240 CFM. In this data hall, that corresponds to less than 1 ACH. Even with two units (480 CFM combined), the result is still only ~1.7–1.8 ACH, far below the 6–10 ACH requir

Design limitations further reduce effectiveness. The unit uses thin pads (~1 in.) at high face velocity, which yield an EBCT of ~0.03 s. Effective sulfur control requires pellet beds 2–4 in.

deep, lower face velocities (50–100 fpm), and EBCT >0.1–0.3 s. The units are carbon-only, without KMnO₄ media, and cannot oxidize reduced sulfur gases (H₂S/SO₂). Carbon-only efficiency also drops at high humidity. Placement as a free-standing unit rather than in the main return path means much of the room air never passes through the filter.

Ceiling height further weakens the coverage claim. The manual advertises 1,000 sq ft at 8 ft ceilings (~8,000 ft³). At 11 ft, that is ~727 sq ft equivalent; at 12 ft, only ~667 sq ft. In a 1,400 sq ft, 12 ft ceiling room (~16,800 ft³), actual coverage is less than half of what is required, even with multiple units. Consumer devices advertise coverage in square feet, but engineering design requires sizing to cubic feet and ACH. Proper industrial HEGA units are specified in airflow

(CFM) and bed geometry, not just floor area. They deliver 6–10 ACH, use pellet beds 2–4 in.

deep, maintain EBCT >0.1 s, and include KMnO₄ media.

Filtration Unit Sizing Analysis

Observed Unit

Inspection and historical records confirm that a small stand-alone gas-phase unit, similar to the

Aerus Beyond Guardian Air class, was installed. This unit is sized for residential or single-room use (≈1,000 ft² maximum coverage), not for the ~1,400 ft², ~14,000 ft³ volume of the Sheriff’s data hall.

Airflow and ACH Analysis

• Data hall volume: ~14,000 ft³ (1,400 ft² × 10 ft height assumed).

• Target ACH for corrosion control: 6–10 air changes per hour.

o 6 ACH ⇒ ~1,400 CFM.

o 10 ACH ⇒ ~2,333 CFM.

• Observed small unit flow: ~200–400 CFM ⇒ <2 ACH. This is an order of magnitude below requirements.

Sorbent Bed Depth and Contact Time

• Industrial HEGA units use 2–4 in. pellet beds at 50–100 fpm face velocity to secure adequate Empty Bed Contact Time (EBCT) for H₂S/SO₂.

• Small chassis units use ¼–1 in. pads at high velocity (often >300 fpm), reducing EBCT to milliseconds—insufficient for low-ppb sulfur capture.

Application

• Report findings: carbon-only media, no KMnO₄ oxidation stage.

• At Florida humidity, carbon alone underperforms and can desorb previously captured

VOCs. KMnO₄ blend is required for reduced sulfur gases.

HEGA Placement

• Manufacturer calls for clearance; inspection notes show the unit was blocking airflow.

Short-circuiting further reduces the percentage of room air exposed to the sorbent.

Evidence of Under-Control

• Latest ISA coupon: Ag 1,824 Å (ISA G3 Harsh) vs. Cu 447 Å (G2 Moderate). This silver-dominant corrosion is classic evidence of sulfur breakthrough when EBCT and

ACH are inadequate.

Engineering Conclusion

Even if “silver-targeted” media were installed, the observed class of unit lacks the CFM capacity, bed depth, velocity control, and placement needed for a data center of this size. Industrial HEGA systems with pellet-style KMnO₄/carbon media and high-capacity fans are required.:

• Silver corrosion nearly quadrupled, indicating sulfur gases are not being removed.

• Copper slowed slightly, confirming the imbalance in filtration effectiveness.

• The facility has transitioned from balanced Cu/Ag corrosion to silver-dominant attack, which directly threatens PCB finishes, connectors, and immersion silver plating.

Conclusion electronics environment:

1. Airflow deficiency: Even with two units, the space remains under 2 ACH versus the 6–10

ACH target.

2. EBCT deficiency: Thin pad design yields EBCT too short for sulfur removal.

3. Media deficiency: Carbon-only cannot arrest H₂S/SO₂; KMnO₄ media is required.

4. Placement deficiency: Free-standing units do not integrate with room airflow.

Therefore, even with multiple consumer units, the data center remains exposed to corrosive sulfur attack, as proven by ISA coupon silver film growth (1,824 Å, G3 Harsh). Only an industrial HEGA system with KMnO₄ media, 1,600–2,800 CFM airflow, engineered bed depth and EBCT, and integration into the recirculation path can adequately mitigate the corrosion risk.

The Beyond Guardian Air (F159E) is undersized and chemically unsuited for a corrosive, critical electronics environment:

1. Airflow deficiency: Even with two units, the space remains under 2 ACH versus the 6–10

ACH target.

2. EBCT deficiency: Thin pad design yields EBCT too short for sulfur removal.

3. Media deficiency: Carbon-only cannot arrest H₂S/SO₂; KMnO₄ media is required.

4. Placement deficiency: Free-standing units do not integrate with room airflow.

industrial HEGA system with KMnO₄ media, 1,600–2,800 CFM airflow, engineered bed depth and EBCT, and integration into the recirculation path can adequately mitigate the corrosion risk

The environment has worsened since 2024, particularly in terms of silver corrosion. The undersized HEGA unit installed was not adequate, and its lack of potassium permanganate media is the primary reason sulfur-driven silver attack has intensified. External conditions (landfill H₂S, wastewater SO₂, airport emissions) remain constant or worsened due to hotter and more humid weather patterns.

• The data center is now classified as ISA G3 – Harsh.

• PCB finishes, immersion silver, and silver-based connectors are at high risk of failure.

• Internal contributors (steam humidifier with contaminated feedwater) are compounding the external problem.

Recommendations

1. Replace HEGA Unit with Proper Industrial Model o Large-capacity unit with pellet-bed media.

o Blend of activated carbon + potassium permanganate (KMnO₄) to specifically remove H₂S and SO₂.

2. Tighten Humidity Control o Adjust setpoints to ≤50–55% RH.

o Place dehumidification ahead of gas-phase filtration.

3. Mitigate Steam Humidification Risk o Replace with RO/DI supply or disable until corrosion trends improve.

4. Rejuvenation and Hardening o Apply HVAC Armor coil coating to reduce future attack.

o Replace corroded UPS batteries and affected components.

5. Ongoing Monitoring o Monthly ISA coupon placement for 90 days.

o Quarterly TO-15 air testing by an independent accredited laboratory.

o Establish trendline reporting on silver corrosion rates.

References

• ISA Standard ANSI/ISA-71.04-2013 – Environmental Conditions for Process

Measurement and Control Systems: Airborne Contaminants.

• ASHRAE TC 9.9 White Paper – 2011 Thermal Guidelines for Data Processing

Environments.

• Independent Accredited Laboratory Reports (SEM/EDS Debris Analysis, TO-15 Gas

Analysis, 2025)【52†source】.

• Sterile Environment Technologies, Inc. – Previous Data Center Consulting Report

(2024).