DEV Community: Cryolab Global

The Failure Is Always at the Interface: A Cold Chain Post Mortem Pattern

Cryolab Global — Fri, 25 Sep 2026 12:46:10 +0000

Frozen biological shipments fail at handovers, not in transit. The vessel is nearly always adequate. The failures sit in the gaps: laboratory to courier, courier to customs, customs to receiving laboratory.

If that sounds familiar, it is the same shape as most distributed systems incidents. The components work. The contract between them is undefined.

Three interface failures worth naming

1. Unverified state at handoff. A dry shipper holds nitrogen absorbed into a porous matrix. Saturation is not visually observable. The fix is a measurement, not an inspection: weigh before and after charging, record both. Weight is the only practical proxy for cold mass actually in the vessel.

2. A published figure treated as a guarantee. Hold times are measured upright, unopened, at moderate ambient temperature. Real shipments satisfy none of those reliably. Treat the published figure as a ceiling measured under stated conditions, and size margin against the delay you did not plan for.

3. Two specifications, constantly conflated. Under the IATA Dangerous Goods Regulations the shipper travels as not restricted under Special Provision A152, provided the packaging cannot build pressure or release nitrogen in any orientation and the air waybill says so. Packing Instruction 650 governs the biological samples inside, classified UN 3373 Category B. A152 covers the vessel. PI 650 covers the samples. A dry shipper is not exempt from PI 650; PI 650 was never about the nitrogen.

The interface contract that actually works

A named person at each end, not a department. Confirmation that the receiving side has capacity before dispatch. Tracking and expected arrival sent directly rather than via the courier's notification. Hold time and its expiry date stated in writing, so a delay becomes a decision rather than a discovery. And a weight recorded on arrival, which is the only feedback loop that improves the next shipment.

Full write up: https://cryolab.co.uk/sample-shipper-frozen-shipments/

When a Single Number Is the Wrong Specification: Temperature Gradients in Cryogenic Storage

Cryolab Global — Fri, 25 Sep 2026 11:48:41 +0000

A specification failure worth borrowing from laboratory cryogenics.

Cryogenic storage is universally specified as minus 196 degrees Celsius. That figure is correct for exactly one configuration: samples immersed in liquid nitrogen, which boils at minus 195.8 at atmospheric pressure and holds the whole canister at its boiling point.

Move to vapour phase storage, where samples sit above the liquid in cold gas, and the single number stops being a number at all. Sample temperature becomes a function of height above the liquid, the liquid level, how often the lid is opened, and the vessel design.

Bottom of the canister sits near the liquid temperature. The top sits warmer. How much warmer is not a constant, it is a function of state that drifts.

Specify against the tail

The sample that matters is the one at the top of a canister in a vessel running low. Nothing about it looks different from the outside. Specify against the worst position in the distribution, not the mean or the best case.

Monitor, do not sample

The ESHRE recommendations on good practice in the IVF laboratory (Human Reproduction 41(8), 1245) expect continuous monitoring with alarms that detect and log out of range temperature or level. Continuous and logged, not sampled. A once-daily manual check is a point observation of a variable that moves between observations, which is the same reason nobody monitors production on a once-daily cron.

Find the leading indicator

The thing holding the temperature is a vacuum between two vessel walls, maintained by a getter with finite capacity. It degrades gradually, and the observable is nitrogen consumption. A rising trend at unchanged workload is the leading indicator. The visible symptom, frost on the outer shell, is a terminal one.

Full write up: https://cryolab.co.uk/liquid-nitrogen-storage-temperature/

Forecasting Consumption From the Wrong Variable: A Cryogenic Storage Example

Cryolab Global — Thu, 24 Sep 2026 13:53:38 +0000

There is a forecasting pattern worth stealing from an unusual domain: laboratory liquid nitrogen supply.

The naive model is demand driven. Samples frozen per month, multiplied by nitrogen per operation, equals consumption. It fails badly.

The real model is dominated by an idle term. A vacuum insulated vessel loses nitrogen continuously whether anyone touches it or not. That baseline, the static evaporation rate, is measured with the vessel closed and undisturbed under defined conditions. Actual consumption is always higher, because every lid lift and every warm object introduced adds heat the measurement deliberately excludes.

consumption = static_boil_off(vessel, ambient) + handling_overhead(access_events)

In a busy store the first term dominates. Halving throughput barely moves the bill, which is exactly the counterintuitive result you get whenever a system's cost is driven by a standing rate rather than a transactional one.

Two operational consequences

Size the delivery interval from measured litres per week in situ, not from the throughput figure. Vendor evaporation rates are laboratory conditions, not your room.
Log per vessel, not per site. A rising trend at unchanged workload is the earliest practical indicator of vacuum degradation, and it shows up months before any visible symptom. Aggregate site level logging hides it.

Where else this shape appears

Idle cloud spend, standing energy load, warehouse refrigeration. If your cost is mostly standing, optimising the transactional path is theatre.

Full write up: https://cryolab.co.uk/liquid-nitrogen-supply/

Dependency Order in Cryogenic Storage: Specify the Consumable Chain Before the Hardware

Cryolab Global — Fri, 18 Sep 2026 12:54:00 +0000

Cryogenic sample storage has a dependency chain, and laboratories routinely resolve it in reverse.

The chain:

straw or vial format
  -> sealer (must match the straw)
  -> goblets or visotubes (capacity varies by straw diameter)
  -> canes and sleeves
  -> canister and rack layout
  -> vessel neck diameter and capacity

The vessel sits at the end. It is also the largest cost and the longest lead time, so it is specified first. Everything upstream then has to conform to a decision made before the requirements were known.

Failure mode one: capacity figures without a format

CBS Daisy Goblets are listed at up to 540 0.25 ml straws, 240 0.5 ml straws or 168 high security straws. Three figures, one product, a 3.2x spread. A capacity number quoted without its straw format is not a specification, it is a marketing figure. Always resolve capacity against the format in use.

Failure mode two: unmatched sealer

CBS High Security sperm straws come in 0.3 ml and 0.5 ml and seal with the SYMS or SYMS III sealer. Embryo straws are 0.15 ml and 0.3 ml. Straw and sealer are a matched pair. One sealer per freezing station plus spares, because a shared unit is both a throughput bottleneck and a single point of failure.

Failure mode three: an undocumented second system

Straws and vials are separate organisation systems. Vials interact with racks and boxes, not with canes and goblets. Running both is fine. Running the second one undocumented is how samples are lost.

Failure mode four: labelling with a short horizon

Since the Health and Care Act 2022 took effect on 1 July 2022, the HFEA confirms storage of up to 55 years with reconsent every ten years. Your label has to outlive the staff who applied it, at least one LIMS migration, and possibly a physical relocation. Cryogenic-rated label stock and printers, colour coded tabs at cane level, cryomarkers.

The test

Every straw format in the laboratory should have a documented path to a specific canister position. If that path cannot be written down, the organisation system has not been designed, it has been assumed.

Full write-up: Cryogenic Straws, Canes, Goblets and Sleeves

Dry Shipper Hire vs Purchase: A Decision Framework, Not a Spec Sheet

Cryolab Global — Fri, 18 Sep 2026 10:54:26 +0000

The usual enquiry starts at the spec: which size, which hold time, which model. Wrong entry point. The first branch is own or hire, and it resolves on two inputs: shipment frequency and how much control you need over vessel condition.

First, what the device is

A dry shipper holds liquid nitrogen absorbed into a porous inner matrix, not as free liquid. Samples travel in vapour phase. The vessel can be inverted without spilling because there is no free liquid to escape.

That construction drives the regulatory position, which is worth getting right because it is routinely stated wrongly:

The shipper travels as not restricted under Special Provision A152 of the IATA Dangerous Goods Regulations, provided the packaging cannot build pressure or release nitrogen in any orientation and the air waybill carries the words not restricted together with the special provision number.
The samples inside are governed by Packing Instruction 650, where they are classified as UN 3373 Category B.

A dry shipper is not exempt from PI 650. PI 650 was never about the nitrogen. These two get conflated constantly, including in supplier literature.

Branch: hire

Infrequent shipments. A purchased unit used a few times a year sits empty, still consuming storage space, inspection and charging time.
One off capacity overflow. Needing more than your own unit holds is a hire, not a second purchase.
Capital constraint. Hire moves cost from capex into the shipment, which is easier to approve and easier to recharge to whoever the shipment is for.
Peak cover. Regular baseline plus occasional surges: hire the surge, do not own the peak.

Branch: own

Frequency. The crossover sits lower than expected once staff handling time is counted properly.
Control, the stronger argument. You know your own vessel's history, whether it has been dropped, how it performed last time. A hired unit puts you inside somebody else's inspection regime, and a shipper that has taken a knock can carry a degrading vacuum with no external sign at all.
Readiness. Charging is not instant. Owning means preparing on your timetable, which matters when a transfer date moves.

Hold time: read it as a ceiling

Published figures are measured upright, unopened, at moderate ambient. Real shipments break all three: orientation changes, ambient varies, customs holds happen, and every opening costs hold time that is not recoverable.

Treat the published number as an upper bound under ideal conditions and size the margin against the delay you have not planned for. A hold time quoted without its measurement conditions is not a planning input.

Charging: the actual failure mode

Undercharging is the most common cause of a failed cryogenic transport, and it is entirely preventable.

Saturate the matrix fully. This takes appreciably longer than most protocols allow.
Invert and drain so no free liquid remains. Skipping this leaves free nitrogen in the vessel, changing both its regulatory position and its behaviour in transit.
Weigh before and after, and record both. Weight is the only practical measure of cold mass actually in the vessel, and the only check that catches a matrix that has not taken a full charge.

Safety note worth building into the procedure: BCGA Code of Practice 30 records that asphyxia due to oxygen deficiency is often rapid with no prior warning to the victim. A freshly charged shipper does not share a lift with a person.

Full article: Dry Shipper Hire or Purchase

Specifying a Cryogenic Storage Vessel: A Dependency Chain, Not a Capacity Number

Cryolab Global — Fri, 18 Sep 2026 10:00:56 +0000

If you specify cryogenic storage the way most purchase orders are written, you resolve the dependencies in reverse and then discover the constraints afterwards. Here is the chain as it actually runs.

The dependency order
straw / vial format
└── goblet and visotube compatibility
└── cane and canister layout
└── rack configuration
└── neck diameter
└── vessel

The vessel is the leaf, not the root. It is also the longest lead time and the biggest line item, which is why it gets ordered first and why the chain gets resolved backwards.

Capacity is a derived value, not an input

Headline vial counts assume a format. A goblet rated for a number of 0.25 ml straws holds considerably fewer high security straws. Request the calculation run against your own formats. If a supplier will not do that, the number they quoted is not about your laboratory.

Validation rule worth adopting: every straw format in use must have a documented path to a specific canister position. Anything without one is an unmapped dependency.

Neck diameter is the access versus heat trade

Wider neck: faster retrieval, less time over an open vessel, fewer awkward lifts. Also more heat ingress, higher boil off, shorter hold time, higher consumption for the life of the asset.

Size it on access frequency, not inventory size. A working store opened several times daily and an archive opened monthly are different specifications even at identical sample counts. Specifying archive characteristics for a working store is a silent failure mode. Nothing alarms. The vessel just runs expensive for ten years.

Storage mode is a documented decision

Liquid phase: most stable temperature, negligible gradient through the canister, longest hold time on power loss.

Vapour phase: no direct contact with the nitrogen, removing that cross contamination route. The ASRM and SART Practice Committees note in Fertility and Sterility (2020;114:486-491) that liquid nitrogen is not sterile. The cost is a vertical temperature gradient and much less tolerance for the liquid level falling.

Both are defensible. An inherited or undocumented choice is not, and it is an early question in any inspection.

Static evaporation rate is your run cost

Measured closed and undisturbed under defined conditions, in litres per day or percentage of capacity per day. It drives ten year cost far more than the purchase price. Comparing quotations on capacity alone is comparing on the least informative field.

The manufacturer figure is a controlled baseline, not a prediction. Real consumption is always higher because the measurement deliberately excludes the thing you do all day: opening it. Several multiples of static rate in a busy IVF store is normal operation.

Post installation, stop reading the absolute value and start reading the trend. Baseline your own litres per week, in your own room, at your own workload. A steady rise at unchanged workload is a vacuum question, and it typically surfaces months ahead of visible frost on the outer shell.

Full article

Comparing Cryogenic Vessel Quotations: The Fields That Are Missing From Your Comparison Table

Cryolab Global — Thu, 17 Sep 2026 14:09:32 +0000

If you have two quotations at the same capacity and different prices, the comparison table you built is missing the fields that explain the difference. Here they are.

Field 1: static evaporation rate

A cryogenic vessel boils off continuously regardless of use. Over a ten year ownership window, the difference between a good and a poor evaporation rate routinely exceeds the purchase price delta you are trying to explain.

Two values should be on every quotation: static evaporation rate, and the weekly nitrogen order it implies at your capacity. If they are absent, request them in writing. A refusal to put the rate in writing is itself information about the rate.

Field 2: what is inside the vacuum annulus

The insulation is a vacuum, and vacuums vary in quality. The cost drivers:

number of layers of multilayer insulation
cleanliness of internal materials, which controls outgassing
getter size, the chemically active component that absorbs stray gas molecules

Getter capacity is a direct determinant of how long the vessel holds its performance. None of this is inspectable on delivery, none of it appears in a spec comparison, and all of it resurfaces later as evaporation rate and service life. This is the largest legitimate source of price difference between two vessels that look identical on paper.

Field 3: mechanical duty rating

Neck construction, internal support design and weld quality determine whether the vessel survives a trolley impact. A shock can crack a weld or fracture an internal support with no external evidence, and the vacuum then degrades silently from that moment.

A static store room vessel is a different build from one moved weekly. The cheap quote is often the first being sold for the second.

Field 4: safety devices

BCGA Code of Practice 36 requires at least two independent safety devices on the inner vessel of a pressurised tank, typically a pressure relief valve plus a separate bursting disc, with the outer jacket relief device set to open at not more than 0.5 bar. A quotation without that arrangement is a different product, not a discount.

Above 0.5 bar above atmospheric, the Pressure Systems Safety Regulations 2000 require a written scheme of examination drawn up by a competent person. HSE guidance INDG178 covers the scope. That is a recurring line item attached to the asset.

Field 5: monitoring, included or excluded

The ESHRE recommendations on good practice in the IVF laboratory, in Human Reproduction (41(8), 1245), expect continuous monitoring with alarms that detect and log out of range temperature or LN2 level. Larger vessels often bundle level, temperature and power alarms with a remote output; smaller ones usually do not. Excluding monitoring makes a quotation cheaper and the installation incomplete. The relay to a BMS or autodialler is easy to omit and expensive to omit.

Field 6: everything that ships after the order

Roller bases, trolleys, racks and canisters matched to your actual straw format, a dispensing dewar, a dipstick, and the PPE for the people handling it. Individually small, collectively the reason single line item budgets run short.

Full article: https://cryolab.co.uk/liquid-nitrogen-container-price/

A Regulatory Change Turned Cryogenic Storage From a Queue Into an Accumulator

Cryolab Global — Fri, 11 Sep 2026 14:51:41 +0000

On 1 July 2022 the UK storage limit for eggs, sperm and embryos changed from ten years to 55, with reconsent every ten years. Most of the operational consequences have nothing to do with clinical practice and everything to do with modelling.

The state change
before: bounded retention (10y)
inflow - outflow ~ steady state
capacity planning = rolling problem

after: near-unbounded retention (55y, reconsent @ 10y)
outflow -> trickle
inventory(t) ~ monotonic increase
capacity planning = cumulative problem

A model calibrated on pre 2022 behaviour will systematically underestimate. Sizing against current inventory encodes the old regime.

The physical guarantee, and its boundary
T < ~ -130 C -> glassy state
molecular motion effectively stops
degradative reactions cease
biological age frozen at t_freeze

A blastocyst warmed after twenty years carries the same biological age it had at freezing. The guarantee is conditional on T being maintained.

The failure band is not ambient
DANGER: -130 C < T < -80 C
motion resumed
ice recrystallisation active
sample still visually frozen

This is the counterintuitive part. The risk window sits well below room temperature and produces no observable signal. Damage is cumulative across events, and each event is unlogged unless you log it deliberately.

Implication: retrieval technique is a storage variable. Thirty seconds too long in warm vapour is a write to the sample with no corresponding entry in your records.

Two freezing methods, different risk surfaces
slow controlled-rate:
cooling ~ fraction of a degree per minute
low cryoprotectant concentration
risk profile: equipment-dependent, tolerant of operator variance

vitrification:
very high cryoprotectant concentration, minimal volume
cooling fast enough to bypass crystallisation entirely
risk profile: operator-dependent, intolerant, timescale = seconds

Vitrification has largely displaced slow freezing for oocytes and embryos. Note that it did not remove risk from the system, it moved it from the equipment to the operator, and shortened the window in which an error occurs.

Controlled rate freezers remain in use where sample volume makes vitrification impractical. A laboratory doing both needs both.

An unresolved tradeoff, and why 55 years changes it
open carrier: direct LN2 contact, highest cooling rate,
non-zero cross-contamination exposure
closed carrier: sealed before cooling, slightly slower,
no contact

LN2 is not sterile and can carry viable organisms between samples in a shared vessel. Documented transmission in clinical practice is vanishingly rare, so the trade is a measurable cooling rate reduction against a low probability event.

But probability integrates over retention time:

P(event over store lifetime) grows with retention window
10y store != 50y store for the same per-year rate

Either position is defensible. An undocumented position is not.

Hardware does not last as long as the data
sample retention: up to 55 years
vessel service life: finite, bounded by getter capacity

The getter maintaining the annulus vacuum has finite capacity. When it saturates, performance degrades with no fault state and nothing visibly broken.

So transfer between vessels becomes a scheduled, repeated, documented operation across the retention window, and every transfer is a warming event, which is the exact risk described in the failure band above.

Monitoring

In March 2018 two unrelated fertility clinics in the United States suffered cryogenic storage failures on the same weekend, and thousands of eggs and embryos were lost.

The generalisable lesson is about observability rather than hardware selection. This class of failure is slow and produces no signal until it is far advanced. Without continuous monitoring and alarms, detection depends on a human happening to walk past.

required: continuous level + temperature monitoring
alarm path reaching a person out of hours
O2 depletion monitoring for personnel
escalation procedure, tested rather than filed

Full write up

The Insulation in a Cryogenic Tank Is an Absence, Not a Material

Cryolab Global — Fri, 11 Sep 2026 14:21:57 +0000

If you think of a cryogenic storage vessel as a container with insulation in the wall, most of its failure behaviour will look mysterious. Model it as a vacuum system with a finite consumable in it and the behaviour becomes predictable.

Three heat transfer paths, handled separately
conduction -> minimise solid cross-section (neck, supports, pipework)
convection -> evacuate the annulus, removing the medium
radiation -> multilayer reflective insulation

Convection is not reduced by the vacuum, it is largely eliminated, because the mechanism requires a gas to carry the heat. That is why vacuum loss is not a graceful degradation. You reinstate a transport path the design had removed entirely.

Multilayer insulation and the constraint that makes it work

MLI is alternating reflective film and low conductivity spacer. The reflective layers handle radiation. The spacers exist to prevent layer to layer contact.

foil | spacer | foil | spacer | foil ...

Remove the spacers and you have a stack of touching conductors. The spacing constraint is the design, not a manufacturing convenience.

Bulk tanks commonly substitute perlite under vacuum. Functionally equivalent, operationally different: a perlite annulus is not a field serviceable component.

The getter is a consumable with no telemetry

Sealed vacuum is not a steady state.

sources: outgassing from annulus materials (slow, years)
H2 and He permeation through steel
sink: getter (chemically active), + adsorbent at cryo temp

The getter has finite capacity. When the sink saturates, annulus pressure rises monotonically and thermal performance degrades with it.

This is the actual reason these vessels have a service life. It is also a failure mode with no direct instrumentation, no error state and no visible fault. You infer it from consumption.

The metric to specify on
NER = static (normal) evaporation rate
= litres/day, or % capacity/day
measured closed and undisturbed, defined conditions

Capacity and price dominate most procurement conversations. NER is the one that determines ten year cost, and two vessels at the same price can differ substantially on it.

Important caveat: NER is a baseline under laboratory conditions, not a prediction for your site. Access events add heat that the measurement excludes.

real_consumption = NER + f(lid_lifts, retrievals, warm_mass_introduced)

A busy store running at several multiples of NER is nominal, not faulty. Which means the absolute number tells you little after installation. The derivative is what you watch.

baseline your own litres/week, then alert on drift at constant workload
Observable failure signatures
Frost or condensation on the outer shell where previously dry, typically first at the base or a support point. Any cold spot on an ambient temperature surface indicates a conduction path through the annulus.
Consumption trending upward over months at unchanged workload.
Hold time below the same vessel's previous year performance.
Behaviour change following mechanical shock. Impact can crack a weld or fracture an internal support with no external evidence.

A degrading vacuum does not self correct and the rate increases once initiated. Treat detection as a decommission trigger, not a monitoring state.

Applicable standards

BS EN 13458 series, three parts: fundamental requirements; design, fabrication, inspection and testing; operational requirements.

BCGA Code of Practice 36 for storage at users premises, including safety device arrangement and the written scheme of examination requirement under the Pressure Systems Safety Regulations 2000 above 0.5 bar gauge.

Open neck dewars vented to atmosphere fall outside the pressure regime but remain vacuum insulated vessels, so the degradation model above applies unchanged.

Full write up: https://cryolab.co.uk/vacuum-insulated-cryogenic-storage-tanks/

Working Out Whether Your Cryogenic Store Can Actually Kill You

Cryolab Global — Thu, 10 Sep 2026 13:58:29 +0000

Most safety writing tells you to be careful. This one gives you the calculation, because the number is more persuasive than the advice.

The expansion ratio

Liquid nitrogen expands approximately 683 to 1 by volume as it warms from its boiling point to room temperature. This figure is published in BCGA Code of Practice 30.

gas_volume_litres = liquid_litres * 683
room_volume_m3 = length * width * height
gas_volume_m3 = gas_volume_litres / 1000
A worked example
Vessel: 175 L
Room: 4 m x 5 m x 3 m

gas_volume = 175 * 683 = 119,525 L
= ~120 m3
room_volume = 4 * 5 * 3 = 60 m3

ratio = 120 / 60 = 2.0

The vessel holds twice the room's own volume in gas.

What that does to oxygen

Air is about 20.9 per cent oxygen. Displacing room air with an inert gas dilutes it proportionally. In the example above, a total release drops oxygen to roughly 7 per cent.

Note the assumption. This treats the room as sealed and perfectly mixed. Real rooms leak and real releases are gradual, so treat it as a bounding case, not a forecast. Bounding cases are how you size ventilation.

Threshold values worth hardcoding

HSE recommends workplace oxygen remains above 19.5 per cent.

The University of St Andrews uses two trigger points: 19.5 per cent for urgent investigation, 18 per cent for immediate evacuation. If you need defensible numbers for a procedure document, use published ones rather than deriving your own.

The part with no sensor

Nitrogen is inert and already makes up most of the atmosphere. Humans have no receptor for it and no receptor for oxygen deficiency. CP30 states that asphyxia due to oxygen deficiency is often rapid with no prior warning to the victim.

In systems terms, you are operating a hazard with no built-in feedback path. The instrumentation is not a nicety, it is the only feedback path that exists.

Alarm placement is an interface problem

Put the alarm indicator outside the door. A monitor that only alerts inside the room notifies the person who is already exposed. Moving the indicator to the outside face of the door changes who receives the signal and when.

Same class of problem: vessels travel in lifts alone. A lift is a small unventilated volume, and a person should not be a passenger in it.

A failure mode you can design out

Open neck storage dewars are not pressure vessels. The closure is a loose insulating plug and the neck vents to atmosphere.

Moisture from the room can freeze in the neck. CP30: an ice plug can form a seal that prevents the release of gas, leading to a rise in internal pressure. Terminal state is vessel rupture.

Mitigations: use only the supplied stopper, never a tighter substitute; visually inspect the neck on every access; treat unusual hissing or a swollen neck as an evacuation trigger rather than a maintenance ticket.

Monitoring: pick the leading indicator

ESHRE 2026 good practice recommendations for the IVF laboratory expect cryostorage tanks to be continuously monitored, with alarms detecting and logging out of range temperature or LN2 level.

ASRM's 2020 cryostorage opinion recommends level probes in the liquid over temperature probes in the vapour, because level falls first. Temperature is lagging. Level is leading. Instrument the leading indicator.

Cheapest monitor in the building is still the log book. Record every fill with date, vessel and litres. Increasing fill frequency at constant workload indicates vacuum degradation, and it shows up weeks before a probe reacts.

Longer write up: https://cryolab.co.uk/liquid-nitrogen-tank-safety/

Cryopreservation Procurement as a Dependency Graph

Cryolab Global — Thu, 10 Sep 2026 13:09:27 +0000

Most equipment lists are written as flat inventories. Cryopreservation does not behave like a flat inventory, and that is why items get ordered late. It behaves like a dependency graph, and the usual purchase order traverses it in the wrong direction.

The naive ordering

storage vessel
straws
nitrogen supply

Three nodes, no edges. This is what typically reaches procurement first.

The actual dependency chain
sample format
-> straw type and volume
-> sealer compatibility
-> goblet / visotube selection
-> cane and canister layout
-> vessel neck diameter + rack configuration
-> vessel capacity and model
-> static evaporation rate
-> weekly nitrogen volume
-> refill equipment class

Resolve it top down and every choice is constrained by the one above it. Resolve it bottom up, which is what buying the tank first does, and you have pinned the root of the tree from a leaf.

The classic failure

Straw format chosen after goblets and canes are already purchased. The formats do not fit. Nothing is broken, nothing is faulty, and the only fix is reordering tank furniture.

Invariant worth enforcing: every straw format in the laboratory has a documented path to a specific canister position. If that path cannot be written down, the graph is not resolved.

Capacity is not a fixed requirement, it is a growth curve

A regulatory change in 2022 altered the shape of the curve. Per the HFEA, since 1 July 2022 UK patients can store eggs, sperm and embryos for up to 55 years provided they reconsent every ten years.

old model: inflow - outflow ~ steady state
new model: inflow - (small outflow) = monotonic growth

Capacity plans carried over from the previous regime will underestimate. Sizing on current inventory rather than projected accumulation is how a vessel bought on price is full in year two.

Two fields missing from most quotations
static_evaporation_rate # litres/day or % capacity/day
weekly_nitrogen_order # derived, per vessel

These are the total cost of ownership inputs. Purchase price is the smaller number over a ten year horizon. If the supplier will not put the evaporation rate in writing, treat that as a signal.

Freezing method branches, it does not default
if sample in (embryo, oocyte):
# ASRM 2021 Practice Committee: vitrification as standard of care
procure(kit, carriers, bench_dewar)
elif sample in (sperm, tissue, cells):
procure(controlled_rate_freezer) # size chamber to batch, not to budget

Laboratories doing both need both. This is not redundancy and should be stated explicitly in the business case, because it looks like redundancy to anyone reading the line items.

Monitoring has a requirement most specs half meet

ESHRE's 2026 good practice recommendations state that cryostorage tanks should be continuously monitored with alarm systems that detect and log any out of range temperature or LN2 level.

Two verbs there. Detect and log. Plenty of systems do the first and not the second.

And the notification path is a separate line item that routinely gets cut:

sensor -> local alarm -> [MISSING] -> on call responder

An alarm sounding in an empty laboratory at 02:00 is an unhandled event.

Non negotiable, non technical

BCGA Code of Practice 30 notes that asphyxia due to oxygen deficiency is often rapid with no prior warning, and records the HSE recommendation that workplace oxygen be kept above 19.5 per cent.

Protective equipment and oxygen monitoring are not optional line items regardless of what the budget looks like at sign off.

Full write up: https://cryolab.co.uk/cryopreservation-equipment-procurement-guide/

Cryogenics Below -150 C: The Engineering Behind IVF, Biobanking, and Industrial Applications

Cryolab Global — Fri, 28 Aug 2026 14:47:25 +0000

A technical overview of cryogenic systems, storage vessel physics, and monitoring architecture

Defining the Temperature Regime

Cryogenics covers temperatures below -150°C (-123K). At these temperatures, standard engineering assumptions break down. Most metals become brittle. Standard elastomers fail. Conventional lubricants solidify. The engineering of systems that function reliably at cryogenic temperatures requires specific material selection, design approaches, and testing protocols that differ fundamentally from ambient temperature engineering.

The key cryogenic fluids and their boiling points at atmospheric pressure:

Liquid nitrogen (LN2): -196°C (77K)
Liquid oxygen (LOX): -183°C (90K)
Liquid argon (LAr): -186°C (87K)
Liquid helium (LHe): -269°C (4K)

For biological sample preservation, LN2 is the standard. Its boiling point provides a 66°C margin below the glass transition temperature (-130°C), the threshold at which molecular motion halts and biological stasis is achieved.

Vacuum Insulation Engineering

The vacuum-insulated dewar is the foundational engineering solution for cryogenic storage. Key parameters:

Vacuum level: high-quality cryogenic dewars target residual pressures below 10^-5 mbar. At this vacuum level, conductive and convective heat transfer through the interstitial space is essentially eliminated. Radiant heat transfer remains and is managed through multi-layer insulation (MLI) or reflective surfaces.

Heat leak sources: the primary heat leak pathways in a cryogenic storage vessel are radiation (managed by MLI or polished surfaces), conduction through the neck tube (minimised by long, thin-walled neck designs in thin-neck dewars), and conduction through internal supports (minimised by using low-conductivity materials such as glass-fibre composites).

Vacuum degradation: the performance-limiting factor over time. Outgassing from internal materials, permeation through seals, and micro-leaks progressively reduce vacuum quality. As vacuum degrades, thermal performance degrades. This is measurable through LN2 consumption rate and holding time, but requires active monitoring to detect.

Temperature Stratification in Large Vessels

In large-volume cryogenic storage tanks (350-1800L as in the CryoBank Series), temperature stratification in vapour phase mode is a design consideration. The temperature gradient between the LN2 surface and the top of the vapour space depends on vessel geometry, neck design, and thermal shielding. A vessel that claims -190°C vapour phase performance at the top of the storage space needs to demonstrate this with sensor data across the full height of the vessel.

Monitoring Architecture

For clinical compliance (HFEA Code of Practice in the UK, FDA cGMP for US pharmaceutical applications), cryogenic storage systems require:

Continuous LN2 level measurement (typically using capacitive or differential pressure sensors)
Temperature monitoring at multiple points within the storage space
Remote alarm notification with configurable thresholds
Data logging for audit trail requirements
Backup power for monitoring systems

The CryoBank Series integrates solenoid valve controlled automated LN2 supply, touchscreen interface with data logging, and remote alarm systems with backup battery capability.

Full technical context: https://cryolab.co.uk/cryogenic-technology-complete-guide/