Refrigerant Recovery curriculum

One brings its own compressor and can pull refrigerant out of an appliance that will not run; the other depends on the appliance's own compressor or on a pressure difference. That dependency decides which is usable on a dead unit, which is the situation that matters.

Liquid expands as it warms and a cylinder filled beyond the limit has no vapor space left to absorb it, so the pressure rises against the walls rather than against a gas. Overfilling is therefore a rupture risk rather than merely a handling inconvenience.

Air collects in the high side and cannot condense, so it occupies condenser volume and raises head pressure without contributing any cooling. The check compares the cylinder's pressure against what the chart predicts at its measured temperature — a disagreement is the evidence.

Where liquid and vapor coexist, pressure and temperature move together in a fixed relationship for that refrigerant, which is what makes a gauge into a thermometer. Every diagnostic in this trade rests on that correspondence, which is also why it fails where the refrigerant is not saturated.

Pulling a deep vacuum boils moisture out of the system at room temperature, which is the only practical way to remove water from a sealed assembly. Breaking the vacuum with dry nitrogen between pulls carries the vapor away rather than waiting for it to migrate.

The two machines do opposite jobs — one discharges to atmosphere and the other captures into a cylinder — so using a vacuum pump to empty a charged system vents it. It is one of the clearest regulatory and practical distinctions in the trade.

Opening a system admits moisture, and with the synthetic oils that moisture becomes acid. Replacing the drier whenever the system is opened is the standard response because the old one has already absorbed whatever it is going to absorb.

Recovery time is dominated by the restriction in the path and by the pressure difference across it — hose bore and length, cores left in place, cylinder temperature. A larger machine does not overcome a narrow path, which is why the setup matters more than the equipment.

Changing refrigerant means changing the oil it must circulate with, the fittings that identify it, and the label describing what is now inside. Skipping the label is the omission that hurts the next technician rather than the current job.

Removing refrigerant and removing air and moisture are separate jobs with different machines and different success criteria — a weight or level in one case, a vacuum reading in the other. Conflating them is how a system is opened while still charged.

The sequence matters because valves and hoses have to be opened and closed in an order that keeps liquid out of the compressor and refrigerant out of the atmosphere. Learning the order as a sequence rather than a set of steps is what makes it reliable under time pressure.

Refrigerant left in the hoses is both a loss and a contamination risk for the next job, so clearing the machine and lines is part of finishing rather than an optional tidy-up. The procedure is what keeps small repeated losses from accumulating.

Self-sealing connections close as the hose is removed, which is what keeps the small release at each disconnection within what the standard tolerates. They address the losses that are individually trivial and collectively substantial.

The machine's internal volume holds refrigerant from the previous job, so switching families without clearing it contaminates the next cylinder. That contamination is expensive in a specific way covered later — a mixed cylinder cannot be reclaimed.

Keeping hoses, gauges and machines separate by refrigerant family prevents cross-contamination at its source rather than managing it afterwards. It costs equipment and removes a failure mode that has no cheap remedy once it occurs.

Pushing vapor into the appliance while pulling liquid out moves a large charge far faster than vapor recovery, and it stalls once the liquid is gone. Recognizing the stall is what tells you to switch methods rather than to keep waiting.

Liquid carries far more mass per unit volume than vapor, so recovering liquid first changes the job's duration by an order of magnitude. Starting on vapor with a large charge is the single most common reason a recovery takes all day.

The cylinder's rated service pressure must suit the refrigerant going into it, because a higher-pressure refrigerant in a lower-rated cylinder is a hazard rather than a compromise. The rating is stamped on the cylinder for exactly this check.

Recovering onto a scale makes the fill limit a measured quantity rather than an estimate from the cylinder's feel or its pressure. It is the only method that answers the question the limit actually asks, which is about mass.

Recovery is driven by a pressure difference, so a cold appliance and a warm cylinder work against you. Warming the appliance and chilling the cylinder restores the gradient, which is the difference between a stalled recovery and a finished one.

The level reached while the machine ran was momentary rather than real: refrigerant absorbed in the oil and in the system continues to come out of solution once the pull stops, and the gauge climbs back. Watching for that climb is what distinguishes a finished recovery from one that merely touched the target for a moment.

Isolating the machine and letting the appliance stand confirms the required level actually holds rather than having been touched briefly. It is the same principle as the rebound check, applied as a deliberate step rather than an observation.

Filters, oil and check valves degrade the machine's performance gradually, so a recovery that takes longer than it used to is usually the machine rather than the job. The decline is invisible without a baseline, which is why it goes unnoticed.

Oil comes out with the refrigerant, and it has to be measured, recorded and disposed of as a waste stream rather than discarded. Quantifying it also matters for recharging, since the system needs the oil it lost put back.

Knowing the appliance's charge before starting turns the endpoint into a number rather than a judgment, and it determines which requirements and which equipment apply. Establishing it first is also what prevents the wrong cylinder being selected.

The category is defined by a small charge in a factory-sealed unit, which covers most domestic appliances. Membership in the category decides which recovery levels and which procedures apply, so identifying it is the first step rather than a formality.

The required level depends on whether the appliance's own compressor still runs, with an alternative expressed as a vacuum. Knowing which condition applies before starting is what determines whether the job is a straightforward recovery or a passive one.

A solderless piercing valve leaks over time, so leaving one on a repaired appliance guarantees a future loss. Removing it or replacing it with a brazed fitting is the difference between a repair and a deferred failure.

Passive recovery has no compressor of its own and relies on the appliance's, or on a pressure difference created by heating and chilling. It is slower and it is sometimes the only option available on a small sealed unit.

A factory-sealed unit has no service valves at all, so access must be created before anything can be connected. That single fact is what makes the access-method decision the first real choice on this class of appliance.

A piercing valve is quick and leaves a leak path; a brazed fitting takes longer and is permanent. Choosing between them is a decision about whether the appliance has a future, which is worth making deliberately rather than by habit.

Using the appliance's own compressor moves refrigerant efficiently when it runs, which is why the required level differs by whether it does. When it will not run, the job becomes passive and considerably slower.

The setup creates the pressure difference the missing compressor would have provided — appliance warmed, cylinder chilled, hoses routed so vapor travels the right way. Each element contributes, and omitting one stalls the recovery.

Where the percentage levels are impractical, a vacuum reading provides an equivalent endpoint that can actually be verified with a gauge. It is an alternative rather than an additional requirement, which is the part most often misread.

The percentages describe what certified equipment must achieve, not a measurement the technician takes in the field. Understanding that removes the confusion about how anyone could measure a percentage of an unknown charge.

A small charge through a restricted access point takes longer than the quantity suggests, and a larger machine does not fix a narrow path. Setting the expectation correctly prevents the job being abandoned early as apparently finished.

Repairing and resealing a hermetic system has to be followed by a leak test, because the repair itself is the most likely new leak. Skipping the test means discovering it after the recharge, with the refrigerant already inside.

Refrigerant must be recovered before an appliance is disposed of, and the rules about who may perform that recovery differ from those for servicing. That difference is what allows disposal operations to function without every operator being certified.

The nameplate states the refrigerant and the charge, which decides the cylinder, the equipment and the endpoint. Connecting before reading it is how the wrong refrigerant ends up in a cylinder that then cannot be reclaimed.

Window units, dehumidifiers, vending machines and water coolers differ in where the sealed system is and where access can be made. The variation is physical rather than procedural, which is why familiarity with the layout saves most of the time.

Flammable hydrocarbon charges change the hazard entirely — ignition sources, ventilation and equipment rating all become controls. The refrigerant quantity is small and the consequence of ignoring the difference is not.

The category is defined by the refrigerant's boiling range at atmospheric pressure, which covers the common high-pressure refrigerants. The classification determines the evacuation requirements, so it is identified before the table is consulted.

The required vacuum depends on the appliance class, the charge size and the equipment's vintage together, which is why it is a lookup rather than a single number. Reading one column and ignoring the others is the usual error.

When leaks make a deeper vacuum unachievable, the requirement drops to a level that can actually be reached, because pulling against an open system accomplishes nothing. It is an exception with conditions rather than a general relaxation.

Recovering liquid before vapor moves most of the mass quickly, after which the remainder is vapor work. The sequence continues into the repair: pressure-test with nitrogen, then evacuate, rather than the reverse.

Electronic detection and bubble testing find different leaks, and pressurizing for a test uses regulated nitrogen rather than the system's own compressor. Using the compressor to pressurize introduces air and moisture into the system being tested.

The appliance class determines the required level before any other column is relevant, so reading the table in the wrong order produces a plausible and wrong answer. The class question comes first for that reason.

The requirement changes at a charge-size boundary, which means isolating a component can move the job from one side of the threshold to the other. That is a legitimate technique and it has to be a deliberate decision rather than an accident.

Machines built before and after the standard's change are held to different achievable levels, because the older equipment cannot reach the newer figure. Knowing which you have determines which requirement applies to your work.

Evacuating an isolated component rather than the whole appliance reduces the work, provided the charge has somewhere to go — usually a receiver. Isolating without a destination for the refrigerant is where the plan fails.

The required level must be confirmed before the system is broken into, not assumed from the elapsed time. That verification is the step that separates a compliant job from one that releases refrigerant on opening.

For a repair that will not be followed by opening the system to atmosphere, the requirement is only to reach atmospheric pressure. It is a narrower allowance than it first appears, and the condition attached to it is what limits its use.

Machine capacity has to suit the charge, because a small machine on a large system turns a morning into a day. It is a planning decision made from the nameplate rather than from whatever is on the truck.

Hose bore, hose length and every valve core in the path restrict flow, and they are the largest factor the technician actually controls. Removing cores and using short large-bore hoses is usually worth more than any other change.

Chilling the cylinder lowers its internal pressure, which restores the difference the machine is working against. It is the cheapest way to speed a recovery that has slowed as the cylinder warmed.

Warming the appliance raises vapor pressure and speeds recovery, with limits on how that heat may be applied — a flame on a sealed system is a different hazard entirely. The technique is legitimate and the method is constrained.

Pumping the charge into the receiver isolates a section for service without recovering everything, which saves considerable time. It stops being sufficient once the section to be opened cannot be isolated from the receiver.

Testing with regulated nitrogen after a repair proves the repair before any refrigerant is committed to it. The test pressure is chosen against the appliance rather than set to whatever the regulator will deliver.

Holding the vacuum and watching the rise distinguishes three situations: a leak rises indefinitely, trapped moisture rises and levels off, and a tight dry system holds. The shape of the curve is the diagnosis, not the final number.

A gauge at the pump reads the pump's achievement rather than the system's, and the difference across a restriction is large. Placing it at the system is what makes the reading mean what you think it means.

The charge is weighed in from the nameplate and then confirmed against operating readings, because the weight is the target and the readings are the evidence. Using only one of the two leaves the result unverified.

After a burnout the refrigerant carries acid and debris, so the machine needs protection and the recovered refrigerant needs a specific destination. Recovering it into a clean cylinder for reuse is the mistake this concept exists to prevent.

The defining property is that the refrigerant boils above ordinary room temperature, so the machine operates below atmospheric pressure. Every difference in this chapter — air ingress, purge units, pressure limits — follows from that one fact.

The requirement is expressed as an absolute pressure rather than a gauge vacuum, which is why a manifold gauge cannot verify it. Reading the requirement correctly determines which instrument the job needs.

Pressurizing is limited to a low ceiling because the relief device protects the shell at a modest pressure, and controlled heat or nitrogen is used to reach it. Exceeding the ceiling destroys the protection rather than testing the machine.

A machine running under vacuum draws air in continuously, so a purge unit removes it as a matter of normal operation. That makes purge activity a running leak indicator, which is the subject of the log later in this chapter.

Liquid comes out first and vapor follows using the recovery compressor, with water kept flowing through the tubes. The water is not incidental — it prevents the freezing that splits tubes during the vapor phase.

The disc relieves at the shell's design limit, which is what sets the ceiling on every pressure applied during testing. Knowing the number is knowing why the test pressure is what it is.

Because the inside is below atmospheric pressure, a leak draws air in rather than pushing refrigerant out — so the symptom is rising condensing pressure rather than a falling charge. That inversion catches technicians used to high-pressure systems.

Air raises condensing pressure and cuts capacity long before anything trips, so the machine simply performs worse for a long time. That slow degradation is why the purge log is the diagnostic rather than an alarm.

A high-efficiency unit removes air while losing very little refrigerant per unit of air, which is the whole point of the equipment. Comparing actual loss against that expectation tells you whether the unit is working.

Run time is the leak indicator on a low-pressure machine, because more air entering means more purging. A rising trend in the log identifies a developing leak well before any other symptom appears.

Verifying an absolute-pressure requirement needs an instrument that can read that low, which a manifold gauge cannot. Using the wrong gauge produces a confident reading that means nothing at the pressures involved.

The appliance is brought to atmospheric pressure before opening, and any opening that cannot be isolated is covered. Opening a machine still under vacuum pulls air and moisture into it at the worst possible moment.

Nitrogen use is restricted on these appliances because the refrigerant's boiling point means the pressure behavior differs from a high-pressure system. The boiling-point test is what decides whether the restriction applies.

Controlled heat raises internal pressure without introducing a non-condensable, which is why it is preferred where it is sufficient. The ceiling on how far it may be taken is set by the same relief device as everything else here.

A ruptured disc means the charge went to the room, which is a release, an exposure and a replacement before return to service. It is an event with three separate consequences rather than a component failure.

Water left standing in the tubes during evacuation can freeze and split them, which turns a service job into a major repair. Keeping it flowing or draining it entirely are the two acceptable states — partially drained is the dangerous one.

Freezing develops during recovery as the evaporator temperature falls, so the process is monitored and paused rather than run continuously. The pauses are part of the method, not an interruption of it.

The maximum is set by the relief device, so exceeding it does not test the machine harder — it destroys the protection. That is why the limit is absolute rather than a guideline to be exceeded carefully.

Warming with controlled hot water raises internal pressure without adding a non-condensable, which is why it is preferred where nitrogen is restricted. The hazards are on the water side — tube pressure, flow interruption, and the freezing risk if circulation stops mid-procedure.

An open-drive machine has a motor outside the refrigerant circuit, which means a shaft seal and a separate oil system that a hermetic compressor does not have. Those are the components that leak and that need servicing, so they are where the attention goes.

Two components change the refrigerant's state and two change its pressure, and knowing which does what is the map for every diagnosis. A pressure reading is only interpretable once you know which side of which component it was taken on.

The system divides at the compressor and the metering device, and that division is the frame every service decision is drawn on. Identifying which side a component is on determines what its pressure should be and what a deviation means.

Boiling absorbs heat in one place and condensing releases it in another, which is where the liquid actually sits. That is why recovery targets the liquid first — it is where most of the mass is.

Superheat measures how much the vapor has warmed past boiling, which reports on how the evaporator is being fed. It has to be measured at a specific place, because the same system gives a different answer at a different point.

Subcooling measures how far the liquid has cooled below condensing, which reports on charge level and condenser performance. It is the counterpart measurement to superheat and it diagnoses the other half of the system.

Fixed, thermostatic and electronic devices respond differently to a charge error, which is why the correct charging method depends on which is fitted. Using the wrong technique produces a charge that reads correct and is not.

The compressor type affects how a recovery proceeds — whether it can assist, how it tolerates liquid, and what isolating it involves. It is a planning input rather than a detail, particularly on larger equipment.

Liquid collects in specific vessels, and finding them is where recovery planning starts because that is where the mass is. A recovery that ignores the receiver leaves most of the charge behind.

The three valve positions isolate, open, or provide access, and combining them is how a component is isolated for service. Misunderstanding a position is how a system is opened while still connected to the charge.

A valve core restricts flow substantially, so every core left in the path lengthens recovery and evacuation. Removing them for the duration is one of the largest time savings available on a service job.

Driers are directional and the flow arrow on the shell says which way, because a reversed drier does not filter and may shed. Placement differs between the liquid and suction lines for different purposes.

Latent heat moves far more energy per unit mass than sensible heat, which is why a refrigerant is chosen for where it boils rather than for its temperature. The whole cycle exists to exploit that difference.

Saturation is what connects a gauge pressure to an actual temperature inside the system, which is what makes indirect measurement possible. It holds only where liquid and vapor coexist, which is why superheat and subcooling are separate readings.

Overcharge, undercharge and restriction each produce a characteristic pair of readings rather than a single symptom, so the diagnosis comes from the combination. Reading one gauge alone cannot distinguish them.

A heat pump swaps which coil is the condenser, so the high side moves depending on the mode. Any diagnostic reasoning about sides has to establish the mode first, which is a step that does not exist on a cooling-only system.

The two gauges read the two sides and the center hose is the path to whatever is connected — vacuum pump, cylinder or recovery machine. Knowing what the center hose is doing at each moment is what prevents connecting the wrong thing.

One dial carries both a pressure and a vacuum scale, and misreading which one is in use produces a number that is wrong by more than its magnitude suggests. The confusion is most likely exactly where the two meet.

They remove arithmetic and reading errors and still depend on correct probe placement and correct refrigerant selection. A digital reading that is confidently wrong is harder to doubt than an analogue one.

A manifold gauge has no resolution at deep vacuum, so it cannot distinguish a good vacuum from a poor one — both read as the end of the scale. Only a micron gauge measures in the range the requirement is stated in.

Three units describe the same physical quantity at different scales, and the deep-vacuum requirement is stated in the one with resolution there. Converting between them is routine and the direction is easy to reverse.

The chart turns a gauge reading into a saturation temperature for a specific refrigerant, which is what makes superheat and subcooling calculable. Using the wrong refrigerant's column is a silent error with a plausible-looking result.

A blend has separate columns because it boils and condenses across a temperature range rather than at a point. Choosing the wrong column shifts every derived figure, which is why blends need more care than single-component refrigerants.

Placement and insulation dominate the reading, so a carelessly attached probe shifts the result enough to change the diagnosis. The measurement error is larger than the differences being measured, which is the point.

Checking what is in the appliance before connecting protects your cylinder, because a mixed charge entering it cannot be separated afterwards. Using it is cheap insurance against an expensive and irreversible outcome.

Resolution, calibration and a level surface all affect a weighed charge, and the charge tolerance is often tighter than a careless setup delivers. The scale is an instrument rather than a platform.

Each method finds different leaks — electronic detection is sensitive and prone to false positives, bubbles are definitive and slow, dye persists and needs access. No single method covers the range, which is why confirmation matters.

High sensitivity produces false positives and low sensitivity misses real leaks, so a hit is confirmed by a second method rather than acted on. Chasing an unconfirmed indication is a common way to lose an afternoon.

Displacement determines speed and oil condition determines the achievable depth, and the gap between a pump's rating and what it actually reaches is usually the oil. A pump that will not pull down is far more often dirty than faulty.

Each oil is matched to a refrigerant family because miscibility and lubrication depend on the pairing, so the oil is not an independent choice. Putting the wrong one in produces a system that runs and fails later rather than one that fails immediately.

Synthetic ester oil absorbs moisture from the air within minutes of a container being opened, which is far faster than the mineral oils it replaced. That speed is why containers are resealed immediately and why partial containers are suspect.

Moisture, oil and heat together produce acid, which attacks motor windings and metal surfaces from inside the sealed system. The chain is why moisture control is a lifespan question rather than an efficiency one.

A positive acid test obliges a cleanup before the system goes back together, because reassembling around acid simply repeats the failure. The test is cheap relative to the second compressor.

After a burnout the contamination is distributed through the system, which is why a suction-line drier goes in temporarily and is then removed. Leaving it permanently creates a restriction the system was not designed around.

Open lines, wet nitrogen, an inadequate vacuum and a leaking low-pressure machine are four separate routes, and a job can suffer from more than one. Identifying which applies is what makes the remedy specific.

Ice at the metering device, a color change in the sight glass and intermittent operation are the recognizable signs, and they appear in that rough order of severity. Intermittent operation is the one most often misdiagnosed as electrical.

Comparing cylinder pressure against the chart at a measured temperature proves air is present, because air adds pressure the chart does not predict. It is a definitive test using equipment already on the truck.

Purging air from a cylinder has a correct procedure and a venting rule limiting how it is done, since the purge carries some refrigerant with it. Doing it casually is a release rather than a maintenance step.

Debris, filings and desiccant powder destroy compressors and block metering devices, and filtration belongs where the flow will carry them. Cleanliness during brazing is where most particulate originates.

A mixed cylinder cannot be reclaimed to specification, so the entire quantity becomes a destruction cost rather than a recoverable asset. That is the concrete reason behind every dedication and clearing practice in this curriculum.

Recovered oil is a regulated waste with its own handling and disposal path, which means it cannot simply be discarded with shop waste. Recording quantity is part of that path rather than optional bookkeeping.

Isolating the pump and watching the micron reading rise distinguishes a leak from trapped moisture from a tight system, because the three curves have different shapes. Reading only the final number loses the diagnosis the shape provides.

Breaking the vacuum with dry nitrogen carries moisture out rather than waiting for it to evaporate, and repeating it has diminishing returns. Knowing roughly how many cycles are worthwhile prevents both an inadequate job and an endless one.

Pump oil absorbs what it removes, so it degrades during the job and the pump's achievable vacuum falls with it. Changing it mid-procedure rather than afterwards is what keeps the second half of the evacuation as effective as the first.

Evacuating through large-bore connections on both sides rather than a single restricted port changes the time by a large factor. It is the same restriction principle as recovery, applied to the other half of the job.

Weighing in the nameplate charge is the primary method because it is the only one that does not depend on operating conditions. The readings-based methods confirm it rather than replacing it.

Which measurement to trim by depends on the metering device fitted, because the two devices control different things. Trimming by the wrong one produces a confidently wrong charge.

A blend must be charged as liquid to keep its composition, and it is metered so the compressor is never slugged. The two requirements pull in opposite directions, which is why the technique exists rather than just the rule.

What correct readings look like depends on ambient temperature and on load, so a charge judged against a remembered figure is judged against the wrong conditions. That is the main argument for weighing rather than reading.

Working from most-likely to least-likely locations finds most leaks quickly, where a systematic sweep spends the same effort everywhere. Joints, valve cores and vibration points are where to start.

Test pressures are set by appliance class, and the system's own compressor is never the pressure source because it introduces air and oil. Both halves of that rule are frequently broken together.

A held nitrogen pressure changes with temperature, so a reading taken hours later must be corrected before it is called a leak. Uncorrected, a cold night produces a leak that does not exist.

Proving the repair before refrigerant goes back in is what keeps a failed repair cheap, because the alternative is recovering the charge again. The ordering is the whole content of the rule.

The steps have a required order — recover, change oil, replace elastomers and drier, evacuate, recharge, label — because each depends on the previous one being finished. Reordering them produces a retrofit that has to be partly redone.

Labeling with the refrigerant and oil actually present is for whoever opens the appliance next, who otherwise has to guess or test. It is the cheapest step in the sequence and the most often skipped.

Recording quantities recovered and added, with dates and appliance identification, is both a regulatory obligation and the only record of what a system has been through. Its value to the next technician is the part that is not about compliance.

A disposable cylinder is built for one filling and has no retest regime, so refilling one is operating a pressure vessel outside its design. The distinction is visible on the cylinder and the consequence of ignoring it is a rupture.

The rating on the cylinder must suit the refrigerant going into it, which is a check made before connecting rather than after. A higher-pressure refrigerant in a lower-rated cylinder is a hazard from the moment it is filled.

A refillable cylinder carries a retest date, and an expired one means it may not be filled or transported until it is retested. It is a date on the cylinder rather than a paperwork matter, which makes it checkable at a glance.

The empty weight stamped on the cylinder is what turns a scale reading into net contents, which is what the fill limit is actually about. Without it the limit cannot be applied, only estimated.

The limit is a function of temperature because the liquid expands as it warms, so a cylinder filled to the limit on a cold day is overfilled on a hot one. That is why the rule is stated as a percentage by weight rather than by volume.

A new or previously used cylinder holds air that becomes a non-condensable in whatever goes in next, so it is evacuated before filling. Skipping it introduces the contamination the rest of the practice is designed to avoid.

Caps and collars protect the valve, which is the one part whose failure empties the cylinder abruptly. Between jobs the cap is the whole protection, which is why it is replaced rather than left off for convenience.

Upright, secured and ventilated is the requirement, and a closed vehicle in sun reaches a temperature that matters for the fill limit. The transport conditions interact with the filling decision made earlier.

Marking, labeling and documentation travel with a shipped cylinder because the carrier and the receiver need to know what it is. Omitting them is a shipping violation rather than an administrative lapse.

Preparing refrigerant for reclamation includes the records that accompany it, since the reclaimer must know the source and the expected contents. Those records are also what make a mixed cylinder traceable to a practice rather than a mystery.

An apparently empty cylinder retains a residual quantity that must be recovered before the cylinder is disposed of. Treating an empty cylinder as empty is both a release and a disposal problem.

Temperature, ventilation and segregation all apply to filled cylinders on site, and temperature connects directly back to the fill limit. Storage is therefore part of the same decision rather than a separate housekeeping matter.