PSA Hydrogen Cycle Timing: Adsorption-to-Equalization Pressure Ratio Optimization for 99.99% Purity

Pressure-swing adsorption (PSA) hydrogen units delivering 99.99 percent H2 purity depend on a single setpoint more than any other to balance recovery against purity: the adsorption-to-equalization pressure ratio. Get this ratio 0.05 too high and product purity drops below spec. Get it 0.05 too low and you waste 0.5 percentage points of H2 recovery, which on a 100,000 Nm3/h stream is roughly USD 1.5 million per year in lost feedstock value. This guide explains the engineering behind the ratio, the operational signatures of a mis-set value, and the five adjustments operators can make to recover both yield and purity without changing the adsorbent.

PSA Hydrogen Cycle Fundamentals: What the Equalization Step Actually Does

A typical 4-bed hydrogen PSA runs a six-step cycle per bed: adsorption, pressure equalization, depressurization (or blowdown), purge, repressurization, and standby. The equalization step is the only step where the unit can recover hydrogen without any external energy input. For a deeper comparison of 5A versus 13X in this service, see our PSA hydrogen guide on 5A versus 13X. While one bed is finishing its adsorption phase at full feed pressure (typically 18 to 30 barg), the bed that just finished its blowdown sits near atmospheric pressure. Opening a cross-connect valve between the two beds lets high-pressure gas flow from the high-pressure bed to the low-pressure bed, equalizing their pressures somewhere between feed pressure and atmospheric.

That equalized gas is rich in hydrogen because it was sitting in the void space of the freshly-adsorbed bed, just above the adsorption front. The bed has held back the CO, CO2, CH4, and N2 impurities on the adsorbent, so the gas in the void space is essentially pure H2 with maybe 2 to 5 percent impurity carry-forward. By transferring this gas to a partially-depressurized bed, the unit recovers H2 that would otherwise be lost in the blowdown step. The recovered H2 does useful work on the next cycle: it improves the receiving bed's purity at the start of its adsorption phase, so less feed gas is needed to reach product spec.

The “adsorption-to-equalization pressure ratio” is the gauge pressure of the high-pressure (donor) bed divided by the gauge pressure of the low-pressure (receiver) bed at the instant the cross-connect valve opens. In practice, the receiver bed is not at zero gauge because it has already been through one equalization step (in a 5-bed unit) or because it has just been purged (in a 4-bed unit). The ratio is therefore not infinite but a finite number between 1.05 and 1.40 in real operating plants.

The Pressure-Ratio Trade-Off: Recovery Up, Purity Down

A higher pressure ratio transfers more hydrogen-rich gas from donor to receiver, which directly increases overall H2 recovery. But the donated gas is not pure H2. It contains a small fraction of co-adsorbed CO and CH4 that was sitting near the bed outlet when the donor finished adsorption. As the pressure ratio rises, more of these impurities are pushed forward into the receiver bed, where they have to be re-adsorbed on the next cycle. This consumes working capacity that would otherwise go to capturing product H2.

The relationship is non-linear. At low ratios (1.05 to 1.10), almost all the donated gas is H2 because the front in the donor bed is far from the outlet. Recovery rises roughly linearly with ratio. At moderate ratios (1.10 to 1.20), the front starts to approach the outlet and a small amount of impurity enters the equalization stream. Recovery still rises, but at a slower rate. Above 1.25, the front has reached the outlet and impurity carry-forward grows quickly. Recovery continues to rise because the absolute volume of donated gas is so much larger, but purity falls sharply because the receiver bed cannot fully adsorb the incoming impurity load.

In a typical refinery hydrogen PSA with 5A molecular sieve, the optimum pressure ratio for 99.99 percent product purity sits between 1.10 and 1.18, depending on feed CO2 content. At ratios above 1.25, product purity starts falling below 99.99 percent. At ratios below 1.05, recovery falls below 85 percent and the unit cannot economically justify its operating cost. Operators typically select 1.15 as the design point and trim by 0.02 to 0.05 increments based on actual feed composition.

Why Feed CO2 Content Shifts the Optimum

CO2 is the strongest-adsorbing impurity on 5A molecular sieve under hydrogen PSA conditions. The higher the feed CO2 partial pressure, the more CO2 each bed holds during the adsorption step. This CO2 is sitting near the bed outlet by the end of adsorption because the CO2 front moves faster through the bed than the CO or CH4 fronts. When you equalize, the donated gas carries that outlet-region CO2 forward to the receiver bed.

The practical consequence: as feed CO2 rises, the maximum allowable pressure ratio falls, because any extra donated gas brings extra CO2 contamination. At 0.5 percent CO2 in feed, the design ratio can sit near 1.20 with purity holding 99.992 percent. At 5 percent CO2 in feed, the design ratio must drop to 1.05 to 1.10 to keep product purity inside 99.99 percent. Above 30 percent CO2 in feed (typical of partial oxidation or biogas-derived syngas), top equalization is often skipped entirely and the unit relies on a feed-side recycle compressor to recover H2 from the tail gas.

This is why hydrogen PSAs designed for steam methane reformer (SMR) off-gas (typically 70 to 80 percent H2, 15 to 25 percent CO2, balance CO and CH4) look very different from hydrogen PSAs designed for POX off-gas (typically 60 to 65 percent H2, 30 to 35 percent CO2). SMR-fed units can afford aggressive equalization ratios and run 5-bed cycles with two equalization steps. POX-fed units run simpler 4-bed cycles with single equalization at conservative ratios and shift the recovery burden to a downstream H2 recycle compressor.

Bed Geometry and Equalization Performance

The length-to-diameter ratio (L/D) of the adsorbent bed is the most under-appreciated design parameter in hydrogen PSA. A tall, thin bed (L/D greater than 2.5) develops a sharp adsorption front. The front's mass-transfer zone is compressed into a small fraction of bed length. At the moment equalization begins, the front is sitting near the bed outlet, and the void-space gas above the front is essentially pure H2. Donating this gas gives high equalization gas purity and high H2 recovery.

A short, fat bed (L/D less than 1.0) has a broad front spread over most of the bed. The void-space gas above the front contains more impurity carry-forward because the front has not yet been pushed to the bed outlet by the end of adsorption. Equalization gas purity is lower and the recovery benefit is reduced.

Industrial 5A units therefore use tall beds. The trade-off is pressure drop: tall beds have higher pressure drop across the bed during adsorption, which costs compressor power and limits the upper feed pressure. Most operating units sit at L/D between 1.6 and 2.2, balancing equalization benefit against compression cost. For the full derivation of L/D limits and why 3:1 to 5:1 is the safe operating window, see our adsorber bed aspect ratio guide. Above L/D of 2.5, the incremental recovery gain is small (less than 0.4 percentage points) while pressure drop grows roughly linearly.

Superficial velocity also matters. If gas velocity through the bed during equalization exceeds 0.25 m/s (pressure-drop calculations are covered in our Ergun equation guide), the donated gas does not have time to fully equilibrate with the receiving bed. The result is a lower-quality H2 capture (because the gas does not dwell long enough to redistribute heat and mass) and a faster recovery drop-off at high ratios. The 0.25 m/s figure comes from the dimensionless Peclet number for mass transfer in 5A pellets of typical 1.6 to 2.5 mm bead size.

Equalization Step Duration: When to Stop

Equalization duration is set by bed void volume plus piping volume, not by adsorbent mass. For a 4-bed 5A unit with 20 to 30 m3 of inter-bed piping, the duration that achieves 95 percent pressure equalization is between 8 and 14 seconds. Extending duration beyond this point costs cycle time without any yield improvement.

Three diagnostic tests confirm you have set the right duration:

1. Pressure-versus-time curve: Plot the equalization pressure on the receiving bed against time. The optimum duration is reached when the pressure slope flattens to less than 0.05 barg per second. Adding another 2 to 4 seconds after this point yields less than 1 percent additional pressure equalization.
2. Equalization gas purity: Sample the equalization gas leaving the receiving bed (typically via a side stream tap on the equalization cross-connect line). The H2 content should stabilize between 92 and 98 percent. When the rising purity curve flattens, the step has captured all the recoverable H2.
3. Recovery-versus-duration sweep: Run the cycle at durations of 6, 8, 10, 12, and 14 seconds and plot recovery versus duration. The curve flattens within 2 seconds of the optimum. The flat portion of the curve is the operating window: anywhere inside it gives recovery within 0.2 percentage points of the maximum.

If the duration is set too short (less than 6 seconds for typical piping geometry), the equalization gas leaves the receiving bed at high velocity and the receiving bed's exit gas is more impure than the design value. Product H2 purity will drift downward by 0.005 to 0.010 percentage points per second of under-timing. If the duration is too long (greater than 16 seconds), the cycle slows down, the unit processes less feed per hour, and capacity drops. The cost of over-timing is usually 1 to 2 percent of nameplate capacity per second of excess.

5-Bed Designs with Two Equalization Steps

A 5-bed configuration adds a second equalization step. After the first top equalization between the highest-pressure bed and the lowest-pressure bed, the 5-bed unit performs a second equalization between intermediate-pressure beds. The recovery gain over a 4-bed single-equalization design is 3 to 5 percentage points because each bed donates gas twice and receives gas twice, so the receiving bed has a higher starting H2 inventory at the start of each adsorption step.

The trade-off is valve count and maintenance. A 5-bed 2-equalization unit has 4 to 6 additional pneumatically-actuated valves compared to a 4-bed 1-equalization unit. Each valve is rated for 1.5 to 3 million cycles, and the cycle frequency of the equalization valves is higher than other valves because they switch twice per bed per cycle. Operators planning a 20-year life cycle should budget 4 to 6 valve rebuild events per equalization valve over the unit's operating life.

Diagnosing a Pressure Ratio That Is Set Too High

Three operational signatures indicate that the adsorption-to-equalization pressure ratio is too high. None of these alone is conclusive, but the combination is diagnostic.

Signature 1: Product purity drifts downward over days. If feed composition is steady (verified by upstream analyzer), bed temperature is steady (verified by thermocouples), and feed flow is steady (verified by flow meter), but product H2 purity is dropping at 0.001 to 0.005 percentage points per day, the equalization setpoint has likely drifted upward. This is usually caused by equalization valve seal wear: as the seals age, the valve does not fully close at the end of the step, allowing a small leakage of high-pressure gas to bleed into the receiving bed beyond the equalization window.

Signature 2: Tail gas H2 content falls below 18 percent. Normal tail gas from a 4-bed 1-equalization unit contains 20 to 25 percent H2. If tail gas H2 falls below 18 percent, two scenarios are possible: either the equalization is not transferring enough gas (ratio too low), or the feed gas is bypassing the adsorbent and reaching the tail gas directly (ratio effectively too high due to valve leakage). A quick diagnostic is to check the equalization gas H2 content on a side stream tap. If equalization gas H2 is also below 90 percent, the ratio is too high.

Signature 3: Receiving-bed temperature profile shifts forward. Most hydrogen PSA beds have multiple thermocouples along the bed length (typically 3 or 4 points: 25 percent, 50 percent, 75 percent, and 90 percent of bed length). The temperature peak (exotherm from adsorption) sits at the location of the mass-transfer front. If the exotherm shifts from the 75 percent point to the 85 or 90 percent point, the front is closer to the bed exit than designed. This means the bed is reaching saturation earlier in the cycle, which is exactly what happens when the equalization ratio is too high and the receiving bed starts its next adsorption with a higher impurity pre-load.

The corrective action is to reduce the equalization pressure ratio in 0.05 increments over three to five cycles and observe whether both purity and recovery improve. If both improve, you are correcting an over-set ratio. If only recovery improves and purity degrades further, the ratio was actually too low and you have moved past the optimum in the wrong direction. The optimum is at the peak of the purity-recovery sum, not the peak of either one individually.

Cold Start vs Steady-State Equalization Setpoints

At cold start, beds are at ambient temperature and clean (no pre-adsorbed species). The dynamic working capacity is at its maximum value, which is typically 8 to 12 percent higher than the steady-state working capacity once the beds reach operating temperature. Operators often raise the equalization pressure ratio to 1.25 to 1.30 during the first 30 minutes of operation to maximize recovery and minimize feed-gas waste during start-up.

The mistake is leaving the cold-start ratio in place as the unit reaches operating temperature. As the beds warm from ambient to the operating range of 35 to 50 degrees C at the outlet, working capacity drops. The receiving bed can no longer absorb the impurity load that the higher equalization ratio is pushing forward. Product H2 purity starts to fall within 90 to 120 minutes of the bed reaching steady-state temperature.

The fix is to tie the equalization setpoint to bed-outlet temperature, not to a fixed clock-time after start. A simple ramp schedule: hold 1.30 for the first 15 minutes, ramp to 1.20 over the next 15 minutes as the bed outlet reaches 30 degrees C, ramp to 1.15 over the next 15 minutes as the outlet reaches 40 degrees C, and hold at 1.10 to 1.15 once the outlet stabilizes at the steady-state value. Without this ramp, the unit will produce off-spec H2 within 2 hours of start-up.

Some modern control systems automate this ramp using a temperature-compensated equalization setpoint function. The operator enters the steady-state setpoint (e.g., 1.15) and the system multiplies by a temperature factor that smoothly ramps from 1.10 (cold) to 1.00 (steady-state) as bed-outlet temperature climbs. The ramp factor is empirically tuned per unit because it depends on the bed geometry, adsorbent mass, and feed gas composition.

Equalization Duration Drift Over Months

Equalization valves are pneumatically actuated with response times that change as seal O-rings swell or compress, as instrument air pressure drifts, and as solenoid coils age. A 1-second drift on equalization duration is enough to shift recovery by 0.3 percentage points. Without monitoring, this drift accumulates silently and operators attribute the recovery loss to feed composition rather than valve timing drift.

The recommended practice is to log the actual equalization valve open-to-close time on every cycle and alert when it drifts more than 0.5 seconds from baseline. For 5-bed units with two equalizations, the same monitoring must run independently on each equalization valve. A typical alarm scheme: warning at 0.5 seconds drift, action alarm at 1.0 seconds drift, unit shutdown at 2.0 seconds drift.

The data is also useful for predictive maintenance. If a valve's open-to-close time trends upward over weeks, the seal is wearing and the valve needs rebuild at the next planned turnaround. This avoids the more disruptive failure mode of a stuck-open valve mid-cycle, which causes an immediate purity loss and a forced shutdown.

Why Equalization Comes Before Depressurization

The standard hydrogen PSA cycle sequence is: adsorption, top equalization, depressurization, blowdown, purge, repressurization. The equalization step is positioned before depressurization for two reasons.

First, equalization recovers H2 without any external work input. It uses the pressure differential between an adsorbing bed at full feed pressure and a partially-depressurized bed. Sequencing equalization before depressurization preserves the highest-pressure gas in the cycle and routes it to a bed that can still use it. If the sequence were reversed (depressurize first, then equalize), the bed being equalized would be at low pressure and could not drive any useful H2 back into the receiving bed.

Second, the receiving bed is at its cleanest state immediately after its own blowdown. The bed has been swept by purge gas, the adsorbent surface is largely free of impurities, and the void space contains purge gas (mostly H2, with some impurities). At this state, the receiving bed has maximum working capacity to capture the donated gas. If equalization happened later in the sequence (after a partial repressurization, for example), the receiving bed would already be partially loaded and could not absorb as much H2 from the donor.

This sequencing rule applies to all PSA designs, not just hydrogen. It is the most basic cycle-design constraint and is rarely violated in commercial units. The only design where it is broken is in rapid-cycle PSA (cycle times below 30 seconds) where the equalization step is sometimes skipped entirely because there is not enough time to perform it within the cycle window.

Cycle Time Allocation Across Steps

For a 4-bed hydrogen PSA with a 300-second cycle, the time allocation across steps is roughly: adsorption 50 percent, equalization 5 percent, depressurization 10 percent, blowdown 5 percent, purge 15 percent, repressurization 10 percent, standby 5 percent. The equalization step is the shortest of the major steps, but its impact on recovery and purity is among the largest per second.

5A Molecular Sieve Grade Selection for PSA Hydrogen

Not all 5A molecular sieve grades are equivalent for PSA hydrogen service. Three specifications drive the cycle performance: bead size, attrition index, and bulk density.

Bead size determines the balance between pressure drop and mass-transfer rate. Smaller beads (1.6 to 2.5 mm) give faster mass transfer and sharper fronts, which improves the equalization gas purity. They also give higher pressure drop, which limits the maximum allowable feed gas velocity. Larger beads (3.0 to 4.0 mm) give lower pressure drop but broader fronts, which means the equalization gas carries more impurity carry-forward. For tall hydrogen PSA beds (L/D above 2.0), 1.6 to 2.5 mm beads are typical. For shorter beds (L/D below 1.5), 2.5 to 3.5 mm beads are common.

Attrition index matters because hydrogen PSA units cycle beds thousands of times per day, and dust generation from bead-to-bead contact and bead-to-wall contact contaminates the bottom of the bed and downstream piping. A typical attrition index specification for hydrogen PSA 5A is below 0.05 percent by weight after 1000 rotations in a standard test drum. Higher attrition grades (0.1 to 0.3 percent) generate visible dust within 12 months of operation and force more frequent filter changes.

Bulk density determines how much adsorbent fits in a given vessel volume. Higher bulk density means more adsorbent per cubic metre, which directly translates to higher working capacity per vessel. For hydrogen PSA, bulk densities above 720 g/L are preferred. Lower bulk density grades (650 to 700 g/L) underperform in the same vessel and force larger vessels for the same throughput.

At AluminaWorld, we supply 5A molecular sieve beads in 1.6 to 2.5 mm and 3.0 to 3.5 mm grades, with attrition index below 0.05 percent and bulk density above 730 g/L. Each lot ships with a Certificate of Analysis showing bead size distribution, attrition, bulk density, and static H2 adsorption capacity at 100 kPa and 25 degrees C. For high-purity hydrogen plants with 99.99 percent target purity, the 1.6 to 2.5 mm grade is the typical recommendation because it gives the sharpest front and the cleanest equalization gas. For a broader PSA energy trade-off discussion across heatless and heated-purge designs, see our PSA sizing comparison.

Process Integration with Downstream H2 Compression

Hydrogen PSA design does not exist in isolation. The outlet pressure of the PSA sets the suction pressure of the downstream H2 compressor, which sets the recycle loop operating point. Equalization pressure ratio interacts with this loop because any H2 left in the tail gas after blowdown can be recovered by recycling the tail gas through the feed compressor.

In plants where the downstream compressor is sized for the full H2 product stream (typical of refinery hydrogen networks), the tail gas is sent to a fuel-gas header or to a steam reformer as supplementary fuel. In these cases, the operator's economic priority is to maximize H2 recovery at the PSA itself, even at the cost of slightly higher pressure ratio. The downstream compression cost is fixed.

In plants where the downstream compressor is sized for a partial recycle stream (typical of merchant hydrogen plants or specialty gas plants), the operator has a choice: maximize PSA recovery (lower tail gas flow but higher H2 content in tail gas) or minimize tail gas H2 content (lower recovery but lower recycle compression cost). The optimum depends on the relative cost of feed gas versus compression power. In regions with cheap natural gas feed and expensive electricity, the optimum leans toward maximum PSA recovery. In regions with expensive feed and cheap electricity, the optimum leans toward minimum tail gas.

For a typical Asian merchant hydrogen plant with USD 6 to 10 per MMBtu natural gas and USD 0.06 to 0.10 per kWh electricity, the optimum sits at 88 to 90 percent PSA recovery. This corresponds to an adsorption-to-equalization pressure ratio of 1.10 to 1.15 on 5A molecular sieve with 1 to 3 percent CO2 feed.

Talk to a Hydrogen PSA Engineer

If you are designing a new hydrogen PSA unit or troubleshooting recovery loss on an existing one, the equalization pressure ratio and duration are the first two parameters to check. Most operators optimize these once at commissioning and never revisit them, but feed composition, adsorbent age, and valve timing all drift over months and years. Re-baselining the optimum every 6 to 12 months is a low-cost way to recover 1 to 3 percentage points of H2 recovery without capital investment. For a worked example of regeneration heat duty calculation on a 1000 kg adsorbent dryer, see our thermal regeneration heat duty guide.

For help selecting the right 5A molecular sieve grade for your hydrogen PSA, requesting a sample for pilot testing, or reviewing your cycle timing parameters, contact our applications team. We supply 5A beads in 1.6 to 2.5 mm and 3.0 to 3.5 mm grades with COA documentation and have supported refinery hydrogen, merchant hydrogen, and SMR/POX plants across 60+ countries since 2010.