🔬 Laboratory — cleanroom air and sample integrity Live model
You are the lab manager. An ISO 7 aseptic suite of 75 m³ is fed by an air-handling unit through HEPA filters; people shed particles, doors let corridor air in, and the fume hood pulls a fixed exhaust that can steal the room's overpressure. Down the hall an ultra-low freezer keeps four racks of samples at −80 °C. Every five minutes the model balances the particles in the room, the airflow through the door gaps and the heat flowing into the freezer. Educational simulation only — not GMP, compliance or clinical advice.
What you will learn
Why a well-mixed room settles at C = G/Q and recovers as e^(−N·t): people set the level, air changes set the speed.
How a pressure cascade really works: the offset between supply and exhaust, the square law through door gaps, and what open doors and a fixed hood exhaust do to it.
How fast a failed −80 °C freezer warms (Newton's law), and why the samples are saved by acting at the alarm, not at the limit.
Simulator
Time 0 min
✓Rack at or below −70 °C
!Rack between −70 and −60 °C
⚠Rack above −60 °C (excursion)
◍Person in a cleanroom coverall
○Person in a lab coat
•Airborne particles (log scale)
❄Compressor running / hood in use
✕Compressor stopped / hood in standby
Controls
Supply air through the ceiling HEPA filters. More air dilutes and clears particles faster, but fan power rises with the cube of the flow. A degraded unit cannot deliver what you ask.
Everyone sheds particles; everyone who walks in or out opens the door. Fewer people also means less work done.
Cleanroom coveralls roughly halve the ≥0.5 µm particles a person sheds compared with a lab coat over indoor clothing. Changing takes about 10 minutes.
One airlock door at a time and short openings: about 10 s and 0.5 m³ of corridor air per opening instead of 40 s and 2 m³. Waiting at the airlock costs a little time.
The exhaust follows the supply minus this offset, which leaks out through the door gaps and holds the room above the corridor. The exhaust can never fall below the hood and safety cabinet's own exhaust.
Cuts the fixed exhaust from 1100 to 500 m³/h, freeing air for the room's overpressure. Work that needs the hood waits.
Stops routine and curiosity openings. Every opening replaces the cabinet air with room air.
Moves racks into the backup freezer's free slots after the call-in time (longer at night). The slots are the building's emergency reserve, released only for a freezer in fault or alarm — a request while it still holds −80 °C is turned down. Each rack spends about two minutes in room air on the way.
Packs remaining freezer racks into insulated chests with dry ice (10 kg per rack, as far as the stock goes). The chest holds −78.5 °C while ice remains.
Indicators
Room class (ISO 14644-1, ≥0.5 µm)
6.86
warning
Pressure difference to the corridor (time average)
12.4Pa
normal
Warmest sample rack
-80.0°C
normal
Sample excursion above −60 °C (rack-minutes)
0min
normal
Particles ≥0.5 µm
256 k/m³
Air changes delivered
20 1/h
Supply fan power
0.45 kW
People in the suite
3
Work done vs. plan
0 % · critical
Pressure difference, doors closed
12.4 Pa
Door openings this period
0
Freezer cabinet temperature
-80.0 °C
Dry ice left
20.0 kg
Racks still in the freezer
4
Trend
Crisis scenarios
Level 1 · Back in class before the aseptic step
It is 08:00. Overnight the suite ran on energy-saving setback at 5 air changes per hour with the hood in standby, and a contractor in street clothes has just finished work inside: the counter reads about 3 million particles per m³. Six people in lab coats are setting up. At 08:30 a three-person team starts an aseptic step that needs the room at ISO 7 or better until 10:00. Keep the team in the room and the fan bill sensible.
Room at ISO 7 or better throughout the step
At least three people in the suite during the step
Average supply fan power ≤ 0.6 kW
Level 2 · Freezer failure at night
It is 21:00 and you are on call. The −80 °C freezer holds four racks of irreplaceable samples. At about 22:00 its compressor stops; the service technician can only come at 07:30. The backup freezer has room for two racks — the building's emergency reserve, released only for a freezer in fault or alarm — and there is 20 kg of dry ice on site, enough for two chests. Whatever you ask for, someone has to come in first: about 30 minutes at night. The security round looks into the freezer from time to time. The lab's SOP says no sample may go above −60 °C, and the group asks for a safety margin: racks should never be warmer than −68 °C.
No rack above −60 °C all night
No excursion minutes at the end
Racks never warmer than −68 °C (safety margin)
Level 3 · Weak air handler, full room
A four-hour hands-on training session has seven people in lab coats in the suite and a steady stream of visitors through the corridor door. Ten minutes in, the supply fan starts failing: it can deliver only 40 % of its capacity (16 air changes per hour) until a repair tomorrow. The fume hood is in use and pulls 1100 m³/h of fixed exhaust. Keep the room at ISO 7, keep at least 10 Pa over the corridor, and get the planned work done.
Room at ISO 7 or better
Average pressure difference ≥ 10 Pa
Planned work done
Basis — the model behind the numbers
Every relation the simulator uses, with its source. Constants marked as assumptions are illustrative calibrations.
ISO 14644-1 sets the class limit for each particle size; one class step is a factor of ten. At 0.5 µm, ISO 7 allows 352 000 particles per cubic metre.
Cn = 10^N·(0.1/D)^2.08 ⇒ N = log10(C) + 2.08·log10(D/0.1); ISO 7 at 0.5 µm = 352 000 /m³, ISO 5 = 3 520 /m³[1]
In a well-mixed room every particle that comes in or is shed is diluted by the supply air and carried out; the level settles where generation equals removal, C = G/Q.
V·dC/dt = G + Q_s·(1−η)·(f·C_out + (1−f)·C) + (Q_in + Q_door)·C_corr − (Q_s + Q_in + Q_door)·C (exact solution per 5-min tick); steady C = G/Q[5][7]
After a disturbance the excess decays exponentially at the air-change rate: a 100-fold drop takes ln(100)/N hours, which is what the ISO 14644-3 recovery test measures.
People are the main source. Measured averages for ≥0.5 µm: about 2.1 million per minute in indoor clothing, about 1.0 million in cleanroom garments; each person and moment varies.
G = Σ people e·exp(0.3·z − 0.045), z ~ N(0,1); e = 35 500 /s (lab coat over indoor clothing; measured for indoor clothing), 17 000 /s (cleanroom coverall), ≥ 0.5 µm[4]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
The offset air leaks out through the door gaps; the pressure needed to push it through grows with the square of the flow. If fixed exhausts take more than the supply leaves over, the room goes negative and draws corridor air in.
Q_off = Q_s − max(E_fixed, Q_s − offset); ΔP = ρ/2·(Q_off/(C_d·A))², C_d = 0.61, A = 0.02 m²; ΔP_avg = ΔP·(1 − time doors open)[6][3][5]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
An open door has no pressure difference; each opening swaps some corridor air into the room. Short openings through an airlock swap far less.
openings ~ Poisson((1·people + traffic)·Δt) + people in/out; per opening 40 s and 2 m³ of corridor air exchanged (10 s, 0.5 m³ with airlock discipline)[3][5]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
Fan power grows with the cube of the airflow: doubling the air changes takes about eight times the power.
P = 0.45 kW·(Q_s / 1500 m³/h)³[8]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
A freezer without its compressor warms like any lumped body: quickly at first, then slower as it nears room temperature. The time constant τ = C/UA sets the clock.
C·dT/dt = UA·(T_amb − T) + Q_door − Q_comp; T(t) = T_amb + (T₀ − T_amb)·e^(−t/τ), τ = C/UA = 20 h; −80 → −60 °C in τ·ln(102/82) ≈ 4.4 h[9]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
Each opening replaces the cold cabinet air with warm, moist room air; its heat and the frost it leaves behind warm the contents.
Q_door = 2·ρ·V_cab·c_p·(T_amb − T) (cabinet air replaced, ×2 for moisture and frost); V_cab = 0.7 m³ → ≈ 1.2 K per opening at −80 °C[9]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
Samples follow their surroundings with a lag; carrying a rack through room air for a couple of minutes costs several degrees.
T_rack ← lag(T_rack, T_surroundings, τ = 1 h); handling: 2 min in room air with τ = 0.5 h (≈ +6 K from −75 °C); excursion = Σ racks minutes above −60 °C[9][11]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
Dry ice sublimes at −78.5 °C and absorbs about 573 kJ per kilogram: the chest stays at that temperature until the ice is gone.
chest at −78.5 °C while ice remains; dm/dt = −(UA_chest·(T_amb + 78.5) + rack heat)/573 kJ/kg[10][11]Assumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
Other operating constants used by the model.
room 75 m³, AHU ≤ 40 air changes/h × health · HEPA 99.95 %, 20 % outdoor air at 35·10⁶ /m³ · corridor 2·10⁶ /m³ · fixed exhaust 1100 m³/h (hood in use) / 500 m³/h (standby) · gowning change 10 min · 1 door trip per person-hour · work × 0.8 with the hood in standby, × 0.95 with airlock discipline · freezer UA 2 W/K, compressor 600 W, alarm −70 °C, SOP limit −60 °C · 4 racks of 15 kJ/K · backup freezer −80 °C, its free slots released only for a freezer in fault or alarm (emergency reserve) · 10 kg dry ice per rack, chest UA 0.25 W/KAssumption: room size, AHU capacity, fixed exhausts, leakage area, door times and exchanged volumes, fan power, freezer insulation and compressor size, rack heat capacity, handling exposure, the emergency-only release of the backup freezer's slots and the −60 °C SOP limit are illustrative values for a generic lab; the physical laws and the published dispersion rates are as cited.
Randomness: a seeded mulberry32 generator; distributions used — uniform, exponential (inverse CDF), normal (Box–Muller), Poisson (Knuth). The seed is shown and shareable.
W. Whyte — Cleanroom Technology: Fundamentals of Design, Testing and Operation, 2nd ed. (well-mixed dilution C = G/Q, decay and recovery, pressure cascades, door openings) — Wiley, 2010
ASHRAE Handbook — Fundamentals, ch. 16 Ventilation and Infiltration (flow through openings Q = C_d·A·√(2ΔP/ρ)) — ASHRAE, 2021
EN 1822-1:2019 High efficiency air filters (EPA, HEPA and ULPA) — Part 1: Classification (H13 ≥ 99.95 % at MPPS) — CEN, 2019
F. P. Incropera, D. P. DeWitt et al. — Fundamentals of Heat and Mass Transfer, 7th ed., §5.1–5.3 lumped capacitance (Newton cooling, τ = ρVc/hA) — Wiley, 2011
ISBER Best Practices: Recommendations for Repositories, 5th ed. (storage equipment monitoring and alarms, backup storage capacity, emergency plans) — International Society for Biological and Environmental Repositories, 2023