Sizing a Heatless (PSA) vs Heated Purge Air Dryer: Purge Loss Percentage and Energy Trade-Off

1. Why Air-Dryer Sizing is a Capex Decision You Re-Live Every Day for 20 Years

A heatless (PSA) regenerative air dryer and a heated-purge blower-purge dryer both deliver the same pressure dew point at the outlet (ISO 8573-1 Class 1 to Class 2 depending on regeneration scheme), but they consume very different amounts of purge gas, very different amounts of electric heat, and they fail in very different modes. Once the dryer is installed and tied into the compressor room, switching from one technology to the other is not a project — it is a capex replacement with new footprints, new piping and a new control narrative. Getting the comparison wrong at the sizing stage costs you 5 to 15 percent of the compressed air energy bill for the next two decades.

The two numbers that drive almost every other design choice are purge loss percentage (the fraction of inlet flow that gets consumed as regeneration purge) and energy per Nm³ of dried air. This guide walks through how each technology actually consumes purge, how to size the desiccant volume on a regeneration cycle that won't channel or co-flow, how to read the vendor cycle-time table correctly, and where the heatless vs heated-purge trade-off bends by inlet temperature, pressure, and intended dew point.

You will see worked examples for 5 Nm³/min, 20 Nm³/min, and 100 Nm³/min inlet flow at 7 barg working pressure, and three worked purge-loss calculations: heatless twin-tower at 15 percent nominal, heatless with mass-flow-controlled cycle at 8 percent, and heated-purge blower-purge at 4 to 6 percent. The formulas are taken from ISO 7183 (Compressed air dryers — specifications and testing), ISO 8573-1 (contaminants and purity classes), ISO 1217 (compressor displacement and FAD definitions), and ASME PTC 28 (gas turbine / adsorption chiller energy accounting, for the energy-balance scaffolding). Vendor-specific numbers in the worked examples are flagged as industry-typical, not factory-specific.

2. The Two Numbers That Drive Every Other Choice

2.1 Purge loss percentage — the silent parasitic load

Purge loss percentage (also called purge-to-product ratio in vendor literature) is the volume of dry, oil-free purge gas bled from the outlet of one tower that is needed to regenerate the other tower, divided by the total inlet flow to the dryer. A heatless twin-tower dryer running at the textbook 5-minute cycle and 4 barg pressure historically consumes 14 to 18 percent of its rated inlet flow as purge — the air you paid the compressor to compress is sent straight to vent after it has done its drying work.

This is not an inefficiency to apologise for; it is a thermodynamic requirement. The heat of adsorption for water on a 4A molecular sieve is roughly 4,200 kJ per kg of water loaded, and at 4 barg the heat capacity of nitrogen at the same molar flow is too low to carry that desorption energy in a five-minute cycle. The vendor adds purge flow until the temperature profile across the bed at the end of regeneration is back to within 5 °C of the inlet purge temperature, and the resulting ratio is what you see on the nameplate.

2.2 Specific power consumption — kWh per Nm³ of dried air

Specific power consumption (SPC) folds together the electric power of the heater (heated-purge only), the blower that re-circulates purge (heated-purge only), the cycle valves' pneumatic actuator air, and the parasitic compressor work of re-supplying the lost purge gas. Industry-typical SPC figures for inlet air at 7 barg, 35 °C inlet, Class 2 pressure dew point (-40 °C):

• Heatless twin-tower at 15 percent purge: 0.18 to 0.22 kWh/Nm³ (most of which is the compressor work of replacing purge)
• Heatless with mass-flow / demand-cycle controller at 8 percent purge: 0.11 to 0.14 kWh/Nm³
• Heated-purge blower-purge at 5 percent purge plus 1.5 kW heater per 20 Nm³/min: 0.09 to 0.12 kWh/Nm³
• Heat-of-compression (HoC) dryer (no purge, uses compressor discharge heat): 0.03 to 0.05 kWh/Nm³ — but only available for oil-free compressors with a usable hot discharge stream

Read those numbers against a 5,000-hour-per-year operating year and an electricity price of 0.10 USD/kWh for an industrial user in most markets. At 100 Nm³/min, the heatless dryer costs USD 162,000 to 198,000 per year in compressor work alone; the heated-purge dryer costs USD 81,000 to 108,000; the HoC dryer costs USD 27,000 to 45,000. That is the trade-off the rest of this guide quantifies.

3. How a Heatless Twin-Tower Dryer Actually Consumes Purge

3.1 The four phases of one half-cycle

One tower of a heatless dryer goes through four phases per cycle, in this order, with the other tower mirror-image out of phase:

1. Online drying — full inlet flow passes top-to-bottom through the loaded desiccant bed; water is adsorbed; outlet dew point meets ISO 8573-1 Class target.
2. Depressurisation — the tower is isolated from inlet, and the working pressure is bled to atmospheric through a muffler or to a buffer tank. Industry-typical depressurisation time is 15 to 30 seconds; the gas lost during this step is approximately 0.5 to 1 percent of the average inlet flow.
3. Regeneration (purge) — a metered fraction of the dried outlet air from the online tower is expanded to near-atmospheric through a fixed purge orifice and blown up-flow through the offline tower. This carries desorbed water out the vent. Typical regeneration time is 4 to 5 minutes; this is the bulk of the purge loss.
4. Repressurisation — the regenerated tower is brought back to line pressure by bleeding dry air from the online tower's outlet. Industry-typical repressurisation time is 15 to 30 seconds; lost gas is again 0.5 to 1 percent of average inlet flow.

3.2 Where the 15 percent number comes from

At 7 barg working pressure, a tower regenerated for 5 minutes with purge at near-atmospheric pressure requires about 1.15 m³ of atmospheric purge gas per litre of bed volume to deliver enough enthalpy to desorb the water loaded during 5 minutes of online operation. The ratio of this purge volume (expanded to working pressure) to the inlet flow during the online phase is what gives you the 14 to 18 percent purge-loss figure.

The dependency is strong: if you raise working pressure from 4 to 10 barg, the purge gas holds proportionally less mass at the same volumetric flow, so the volumetric purge flow has to roughly double to deliver the same energy, and the purge-loss percentage climbs. If you slow the cycle from 5 minutes to 10 minutes, each tower has twice as long to load water but still needs the same enthalpy to regenerate, and the purge-loss percentage again climbs. The 15 percent figure is anchored to a very specific set of operating conditions, not an inherent property of the technology.

3.3 Demand cycle control cuts purge loss to single digits

A demand-cycle controller (also called a mass-flow or dew-point-controlled cycle) holds the tower online until the outlet dew point climbs to a set threshold (e.g., -33 °C for a Class 2 spec at -40 °C target with 7 °C margin), then triggers the switch. Towers that are lightly loaded regenerate faster and need less purge. Industry-typical purge loss under demand control at 7 barg is 6 to 10 percent, and the controller pays for itself in compressor energy in less than 18 months on any dryer above 20 Nm³/min.

4. How a Heated-Purge Blower-Purge Dryer Cuts Purge to 5 Percent

4.1 The blower-purge architecture

A heated-purge dryer adds a third phase between the regeneration and repressurisation steps: a closed-loop blower pulls the hot regeneration gas through an external electric or steam heater, then back up-flow through the offline tower. The closed loop means the only purge loss is what leaks out the vent valve when the desorbed water is expelled; typical blow-through flow is 3 to 8 percent of inlet flow for the bleed step. The blower itself consumes 0.3 to 0.6 kW per Nm³/min of inlet capacity; the heater consumes another 0.5 to 1.5 kW per Nm³/min depending on cycle length and regeneration temperature.

4.2 Why heating changes the energy balance so dramatically

At 200 °C regeneration temperature (typical for heated-purge on 4A or 13X), the gas carries roughly 3.5 times more enthalpy per Nm³ than it would at 30 °C atmospheric purge. The purge gas therefore can desorb the same loaded water at 3 to 4 times lower volumetric flow. Multiply that by the closed-loop recirculation (no fresh inlet flow is consumed except the bleed to remove water), and you end up with purge loss of 4 to 6 percent of inlet flow plus the parasitic cost of the electric heat.

The trade is then a useful energy comparison. At 7 barg, the heatless dryer has to compress the purge gas from atmospheric to 8 bara (factor of 8) before it can be used; the heated-purge dryer only compresses a tiny bleed stream. Even with the electric heater, the total specific power is lower.

5. The Two Sizing Formulas You Cannot Get Wrong

5.1 Desiccant volume — from inlet flow, inlet temperature, and pressure dew point

The mass of water to be adsorbed per cycle depends on inlet flow, inlet temperature, working pressure, and cycle time. A simplified engineering formula (consistent with the methodology in ISO 7183 Annex C) is:

m_H₂O per cycle = Q × t × ρ_air × (x_in − x_out)

where Q is inlet volumetric flow in Nm³/s, t is online cycle time in seconds, ρ_air is inlet air density (≈ 1.20 kg/Nm³ at STP), x_in is inlet absolute humidity in kg water per kg dry air, and x_out is outlet absolute humidity corresponding to your pressure dew point target. For Class 2 (-40 °C PDP) and inlet at 35 °C, 95 percent RH, 7 barg, x_in ≈ 0.030 kg/kg and x_out ≈ 0.00012 kg/kg.

Desiccant volume is then calculated from the equilibrium loading of the desiccant at the regeneration condition (typical values: 4A molecular sieve loads about 10 to 11 percent by weight at 35 °C / 80 percent RH and regenerates to 0.5 to 1 percent residual moisture at 200 °C heatless-purge conditions, or about 3 to 5 percent residual at heatless 150 °C). Industry-typical design uses 60 to 75 percent of equilibrium working capacity as the design loading. The volume is:

V_bed = m_{H₂O per cycle} / (ρ_bulk × W × U)

where ρ_bulk is bulk density of the desiccant (~750 kg/m³ for 4A), W is working weight capacity (kg water per kg desiccant), and U is design utilisation (fraction). At 20 Nm³/min, 35 °C inlet, 7 barg, and 5-minute cycle, industry-typical 4A bed volume is 0.42 to 0.50 m³ per tower; at 100 Nm³/min it scales to 2.1 to 2.5 m³ per tower.

5.2 Tower diameter — L/D, velocity, and pressure drop

Tower diameter is set by superficial velocity through the bed. For 4A molecular sieve in an air-drying service, industry-typical gas velocity is 0.10 to 0.25 m/s at working conditions. Too low and you risk channelling; too high and you crush the beads or blow them out the bed supports. The pressure drop through a clean 4A bed at design velocity is roughly 2 to 4 kPa per metre of bed depth.

The aspect ratio rule (covered in detail in the companion post on bed L/D) is L/D = 2.5 to 3.5 for adsorber vessels in pressure-swing service; smaller L/D leads to axial channelling and incomplete use of the bed, larger L/D leads to high pressure drop and dust fluidisation at the bottom of the bed. Combined with the bed volume above, the tower diameter and shell height follow.

5.3 Quick-reference sizing summary

For a typical industrial compressed-air system at 7 barg, 35 °C inlet, and Class 2 PDP (-40 °C) with a 5-minute heatless cycle, the table below summarises industry-typical bed volume, tower diameter and aspect ratio at three common plant sizes. Use this as a sanity check against any vendor proposal before you commit to the order.

Note that the smallest size has the highest L/D — a 5 Nm³/min dryer is genuinely small and benefits from a tall vessel. The 100 Nm³/min case shows L/D closer to 1; for that size, a vendor may propose two parallel 50 Nm³/min towers to recover a more conservative aspect ratio. Always check whether the vendor is quoting a single tower or twin towers at large sizes — the cycle valves, the purge piping and the depressurisation mufflers all scale with single-tower size.

6. Worked Examples — Purge Loss and Energy Cost at Three Sizes

6.1 Common assumptions for the table

The three worked cases below assume 7 barg working pressure, 35 °C inlet, 95 percent inlet RH, Class 2 (-40 °C PDP) outlet, 5,000 hours/year operation, 0.10 USD/kWh electricity cost, and compressor isentropic efficiency 70 percent (typical for oil-injected screw at 7 barg). Numbers are flagged as industry-typical; actual compressor efficiency in your plant will shift the absolute energy figures by 10 to 20 percent, but the ranking is robust.

6.2 Three sizes side-by-side

The crossover where heated-purge becomes obviously worth the higher capex is roughly 30 Nm³/min: at smaller sizes, the heatless dryer with a demand cycle controller is hard to beat on first cost, and the absolute energy savings don't justify the larger footprint and heater maintenance. Above 50 Nm³/min, the heated-purge dryer is almost always the lower-total-cost option over a 15-year life cycle, especially if you can recover the heater waste heat.

7. Where the Sizing Math Breaks Down

7.1 Very high inlet temperature (> 45 °C)

Compressor rooms with poor ventilation or with a turbo compressor pulling hot ambient air can easily reach 45 to 55 °C inlet. Above 45 °C the heatless cycle needs to either shorten the online time (which raises purge-loss percentage) or accept higher outlet dew point; above 50 °C the heatless dryer often cannot meet Class 2 without an after-cooler pre-treatment. Heated-purge is more tolerant of high inlet temperature because the regeneration temperature can be raised to 220 to 250 °C to compensate.

7.2 Very low inlet pressure (< 4 barg)

At 3 barg working pressure, the volumetric flow through the bed is roughly double what it is at 7 barg for the same mass flow. Pressure drop roughly doubles, and the purge gas at near-atmospheric exit pressure has to be re-compressed through a larger pressure ratio. Both effects push you toward larger desiccant volume and higher specific power. Below 3.5 barg, heated-purge with a separate purge blower is often the only practical option.

7.3 Class 1 PDP (-70 °C) requirements

Class 1 (-70 °C PDP) requires either a desiccant dryer in series (two heatless towers with an inter-cooler) or a single heat-of-compression dryer followed by a desiccant polisher. Heated-purge reaches Class 1 with a single tower if the regeneration temperature is held at 230 to 260 °C and the bed depth is increased by 30 to 40 percent over a Class 2 design. The capex and energy penalty for Class 1 is non-trivial and should be justified by the actual process need; most compressed-air applications outside semiconductor and pharmaceutical do not require it.

8. Cross-Check Questions to Ask the Vendor Before You Sign

Use this list as your sizing-review checklist when a vendor submits a proposal:

• At what working pressure, inlet temperature, and inlet RH is the quoted purge loss percentage measured? Is it at nominal or worst-case inlet conditions?
• What is the cycle time at full load and at 50 percent load? Heatless dryers on demand cycle vary cycle time by 2 to 4× across the load range.
• What is the desiccant volume per tower, and what is the working capacity the design uses for the molecular sieve? A vendor quoting less than 50 percent utilisation of equilibrium capacity is over-sizing the bed (good for safety margin, bad for capex).
• For heated-purge: what is the regeneration temperature set point, and what is the heater kW rating at full load?
• What is the outlet dew point at full load and at 110 percent momentary overload? A heatless dryer that quotes -40 °C PDP at full load may give you -30 °C at overload.
• Is the pressure-drop guarantee based on clean bed or on bed-at-end-of-life? Industry-typical guarantee is "bed at end of design life (typically 3 to 5 years)" — ask for the clean-bed number too.
• What is the noise rating of the depressurisation muffler? Heatless dryers without good mufflers are 85 to 95 dBA at 1 m; a sub-enclosure or remote mounting may be required.
• For retrofit applications: does the proposed unit fit the existing footprint, or do you need new piping and electrical? Heated-purge is usually 30 to 50 percent larger than heatless at the same capacity.

9. Frequently Asked Questions

9.1 What is the difference between heatless and heated-purge desiccant air dryers?

A heatless dryer regenerates one tower while the other dries, using a small fraction of the already-dried outlet air at near-atmospheric pressure as the regeneration purge; no external heat is added. A heated-purge dryer also runs two towers but uses an external electric or steam heater to warm a closed-loop purge stream (typically to 180 to 220 °C) so that much less volumetric purge flow is needed. Heatless dryers are simpler, smaller, and cheaper to buy but consume 10 to 18 percent of inlet flow as purge; heated-purge dryers consume only 4 to 6 percent of inlet flow but cost more up front and need electric heat. Both meet the same pressure dew point class (e.g., ISO 8573-1 Class 1 or Class 2).

9.2 How is purge loss percentage calculated?

Purge loss percentage = (volumetric flow of purge gas consumed per minute, expanded to inlet conditions) / (volumetric inlet flow to the dryer). For a heatless dryer with a fixed purge orifice at 7 barg and 5-minute cycle, industry-typical purge loss is 14 to 18 percent. With a demand-cycle (dew-point-controlled) controller, the purge loss drops to 6 to 10 percent because lightly loaded towers regenerate faster. For a heated-purge blower-purge dryer, purge loss is the bleed fraction needed to remove desorbed water from the closed loop — typically 4 to 6 percent.

9.3 What is the typical payback period for switching from a heatless to a heated-purge dryer?

At 50 Nm³/min inlet flow and 5,000 hours/year, the heated-purge dryer saves roughly USD 45,000 to 60,000 per year in compressor work relative to a 15 percent-purge heatless dryer. The capex premium for heated-purge at this size is in the range of USD 80,000 to 150,000 depending on heater type and instrumentation. Simple payback therefore falls between 1.5 and 3.5 years. Above 100 Nm³/min, simple payback is typically under 2 years. Below 20 Nm³/min, the heated-purge premium rarely justifies itself on energy savings alone.

9.4 Can I retrofit a demand-cycle controller on an existing heatless dryer?

Yes, on most modern heatless dryers that have electronic cycle timers rather than purely pneumatic timers. The retrofit kit is a dew-point sensor at the outlet, a small PLC or dedicated controller, and a software/firmware change to the cycle timer. Industry-typical cost is USD 4,000 to 12,000 installed. The retrofit usually pays back in 12 to 24 months through reduced purge loss alone, before counting any reduced compressor loading. Confirm with the dryer OEM that the controller can be retrofitted to your specific model before ordering.

9.5 How does inlet temperature affect dryer sizing?

Higher inlet temperature means more water is carried per Nm³ of inlet air (warmer air holds more moisture). At 7 barg, raising inlet temperature from 25 °C to 45 °C roughly doubles the absolute humidity, doubling the required desiccant loading per cycle and therefore the bed volume. It also reduces the effectiveness of any after-cooler. Industry-typical practice is to design at the hottest expected inlet temperature for at least 5 percent of operating hours, not at the annual average. In hot climates (Middle East, India, Southeast Asia), this often means over-sizing the dryer by 20 to 40 percent relative to a temperate-climate design.

9.6 What maintenance is required on heatless vs heated-purge dryers?

Heatless dryers have very few moving parts: cycle valves (solenoid or pneumatic), a purge orifice, mufflers, and pressure gauges. Typical maintenance is annual valve rebuild, 3-year desiccant replacement if upstream filtration is good, and quarterly muffler inspection. Heated-purge dryers add a blower (bearings every 2 to 3 years), a heater (element replacement every 5 to 8 years for electric), and more instrumentation (temperature transmitters, pressure transmitters). Annual maintenance cost is therefore typically 1.5 to 2× higher than heatless, but the energy savings usually more than offset it above 30 Nm³/min.

9.7 Does a heat-of-compression (HoC) dryer avoid the purge trade-off?

A heat-of-compression (HoC) dryer uses the hot discharge from an oil-free compressor (typically a centrifugal or oil-free screw) at 150 to 200 °C to regenerate the desiccant, eliminating the need for a separate purge stream. Purge loss is effectively zero (only depressurisation and repressurisation losses, around 1 to 2 percent). However, HoC dryers are limited to oil-free compressors, are larger than heatless dryers at the same capacity, and have higher first cost. They dominate the 50 Nm³/min and above range when an oil-free compressor is already specified for the process (food, pharma, electronics).

9.8 How does working pressure change the purge loss percentage?

Purge loss percentage rises at lower working pressure because the purge gas at near-atmospheric outlet has to be re-compressed through a larger pressure ratio. At 10 barg, a heatless dryer might consume 12 percent purge; at 4 barg, the same dryer might consume 18 to 20 percent. Heated-purge dryers are less sensitive to working pressure because the purge gas is in a closed loop and is not re-compressed from atmospheric. If your plant operates below 5 barg, heated-purge is almost always the lower-energy option.

9.9 What dew point do I actually need?

Most compressed-air applications outside pharmaceutical, semiconductor, and certain analytical-instrument uses do not need Class 1 (-70 °C PDP). Pneumatic tools and general instrumentation typically need Class 3 to Class 4 (-20 to +3 °C PDP). Process air for food contact needs Class 2 (-40 °C PDP) at minimum. If your spec was written by copying from another plant or by defaulting to the lowest number on the vendor quote, you may be paying for a dryer two classes tighter than you need. The energy cost difference between Class 2 and Class 4 is roughly 30 to 50 percent — re-check the spec before you finalise the order.

9.10 What is the typical desiccant lifetime before replacement?

Industry-typical molecular sieve life in a well-maintained air dryer is 3 to 5 years. The degradation mechanisms are: (a) oil carryover from upstream compressor that coats the beads and reduces capacity — fix by installing a coalescing filter, (b) dust attrition from thermal cycling that breaks beads into fines that fluidise and carry over — fix by installing a bottom bed support of the right mesh, and (c) poisoning by compressor lubricant additives that polymerise on the bead surface — fix by switching to a synthetic compressor oil and confirming the upstream filter is sized for the new oil. Sampling the desiccant annually and comparing BET surface area and water loading capacity to the original CoA is the standard health check.

9.11 Can a refrigerated air dryer meet my dew point requirement?

A refrigerated air dryer typically delivers +3 °C PDP at the outlet (ISO 8573-1 Class 4 to Class 5). If your process needs -20 °C PDP or lower, a desiccant dryer is required; refrigerated dryers cannot reach sub-zero pressure dew points economically. However, a refrigerated dryer pre-treatment before a desiccant dryer is a common configuration that reduces the desiccant loading and can shrink the dryer by one size class. The refrigerated pre-treater removes the bulk of the water at low cost, and the desiccant polisher hits the final dew point with much less purge.

9.12 How do I size a dryer for variable load?

If your plant operates at 30 to 100 percent of nameplate flow over the year, you have two choices: (a) oversize the dryer for peak flow and let it cycle slowly at low load (a heatless dryer at 30 percent load still consumes about 10 to 12 percent purge, not the 15 percent at full load), or (b) install a demand-cycle controller that adjusts cycle time based on outlet dew point. The demand-cycle option is almost always the better choice if your dryer is sized for full-load nameplate and spends meaningful time at partial load — the energy savings typically pay for the controller in 12 to 24 months.

10. Next Steps

Send us your compressor room data — inlet flow (Nm³/min), working pressure (barg), inlet temperature (°C), required PDP class (ISO 8573-1), and current annual operating hours — and we will model the purge-loss and energy trade-off for heatless, heatless-with-demand, and heated-purge options side by side. We will also confirm the desiccant volume, the L/D aspect ratio, and the expected cycle time at full and partial load so the proposal you receive from the OEM can be checked against an independent calculation. Engineering contact: barry@aluminaworld.com or WhatsApp +86 133 2522 2240.

Free sample policy: we supply 5 kg of 4A molecular sieve or activated alumina for your bench-scale dryer trial so you can verify the actual equilibrium loading and pressure dew point at your inlet conditions before committing to the full order. MOQ 1 MT production orders, 15 to 25 day lead time, ISO 9001 certified, shipped from Zibo, Shandong to 60+ countries. COA and MSDS included. Documentation language: English, Russian, Arabic, Spanish, Turkish on request.