Scroll / Screw Chiller Guide: 7 Load-Profile Checks (2026)

Updated September 2026 · Informational guide for engineering, procurement and operations teams

Working rule: a rating is only as useful as the load, heat-rejection and included-power boundary attached to it.

What a Scroll / Screw Chiller Actually Means

In a mechanical chiller package, the compressor raises refrigerant pressure so heat can be rejected at the condenser. “Scroll” and “screw” describe that compression method. Neither term specifies the evaporator, controls, refrigerant, pumps, acoustic treatment or operating sequence for the complete plant.

A scroll compressor uses fixed and orbiting scrolls. The orbiting element traps and compresses vapor in pockets. A screw compressor uses intermeshing twin rotors that move and compress vapor along their length. Both are positive-displacement machines. The useful question is how each machine and its controls behave hour by hour at the site.

The screw vs scroll question starts with the compressor mechanism, and the difference between scroll and screw compression is mechanical, not a quality ranking. Rotary screw compressors mesh two screw rotors together as male and female rotors, while a hermetic scroll compressor seals its orbiting mechanism inside one welded shell. Comparing these different types of compressors is a compressor design question, not a single best type compressor answer: whether scroll and screw chillers or screw and scroll chillers suit a site better depends on the load profile covered next, not on the compressor name alone. Either scroll type or screw machine will compress the refrigerant before it reaches a shell and tube evaporator or plate heat exchanger downstream; the differences between the two show up more in part-load staging behavior than in a single scroll vs screw efficiency number.

The rating boundary also matters. AHRI 550/590 describes how a complete water-chilling package is rated; it is not a “bare compressor” test.

Compare candidates using the same entering and leaving chilled-water temperatures, condensing or ambient conditions, flow, fouling assumptions and included power. State whether pumps, fans and towers are included, and whether the number represents one compressor, the chiller package or the whole plant. Without that boundary, “scroll efficiency” and “screw efficiency” are not comparable statements.

Keep the vapor-compression description separate from a compressor-only, chiller-only or whole-plant claim. The AHRI 550/590 performance-rating guidance provides the rating context for the package.

For an engineer reviewing a new factory installation, the practical risk is a specification that looks precise but leaves the evidence boundary unstated. Ask for the factory acceptance test, the calibrated instruments used and the operating scenario represented by the result. That short evidence trail prevents a compressor label from becoming a misleading plant claim.

In a production-line application, ambiguity in the rating creates a procurement risk because the factory test, factory witness record and factory data sheet may describe different boundaries. Add a 2-hour witnessed run and a calibrated trend export to close that evidence gap before approval.

How Scroll and Screw Compression Behave Under Real Loads

A full-load design point can make two options look easy to separate, yet few plants operate there for long. Actual load changes hour by hour with weather, ambient temperature, cleaning cycles and occupancy. Cycling, staging, turndown and the size of the compressor bank can affect annual cost as much as the compressor architecture itself.

Multiple scroll compressors can be staged as load changes. A screw machine may use a slide valve, variable speed or stepped circuits. None guarantees a lower annual cost by itself. A small scroll bank can follow a changing load smoothly, while a poorly sequenced array can run too many machines at low load. A screw with capacity control can also cycle when pressure or heat-rejection conditions move outside its stable range.

In other words, part-load behaviour must be checked with the actual slide-valve and variable-speed sequence, not inferred from the compressor name. The DOE FEMP chiller guidance is a useful benchmark because it keeps full-load and variable-load performance in the same procurement conversation.

Consider a process line that starts a batch every 3 hours: the scenario creates step changes that a steady design-day point cannot show. A trend export, start-count log and commissioning witness test provide better evidence than a claim that one architecture is always safer. The failure mode to avoid is not choosing the “wrong” word; it is choosing without seeing the operating sequence.

Use the architecture as a hypothesis to test the load profile against. In a scroll-versus screw compressor review, request a capacity-control map, a minimum stable operating point, number of starts per hour, minimum unloading limit, oil-management needs, sound levels at expected duty, and the sequence that starts another circuit. These belong in the specification, not in an after-sales discussion.

Start With the Load Profile, Not the Nameplate

The nameplate is an operating boundary, not an operating year. Use at least 12 months of hourly (or finer) data when it is available. For new equipment, a transparent surrogate covering about 2,000 hours of expected operation can be useful. Build it from production hours, weather bins, occupancy, process setpoints and expected diversity, and record every assumption and its likely accuracy.

Keep both the full-load and partial-load cases visible. Even a neat-looking curve is not enough when the project needs sub-hourly data, minimum-flow and leaving-water-temperature limits, and a clear design-day assumption. Record the Design-day basis in the schedule. DOE FEMP guidance is the reference for the full-load/part-load distinction.

Break apart these four loads that are often combined: For the commissioning example below, mark any shift or shutdown longer than 6 hours so it isn’t assumed a minimum-load condition:

• Design-Day Peak: the short period during which the worst probable situation must be protected.
• Typical Operating: where the equipment will spend most of its hours.
• Minimum Stable Load: the lowest load that keeps flow under control with limited cycling.
• Transient or Process Load: batch events such as product removal and cleaning that may need a buffer.

A simple worked example shows why two staged circuits may deserve consideration. Suppose a loop reaches 420 refrigeration tons on a peak afternoon, runs near 240 tons during the day and falls to about 80 tons when the facility is unoccupied. The question is not “which single compressor reaches 420 tons?” It is “which arrangement can manage 80–240 tons efficiently, cover the peak and recover from a standby condition?” The answer may involve staged circuits, a buffer tank, revised controls or a revised design load. If data is incomplete, capture at least 4 representative months, mark batch steps longer than 3 hours and record any documented sensor-offset assumption.

DOE FEMP guidance publishes full- and part-load information because a chiller that looks good at one condition may operate at another for most of the year.

How Part-Load Efficiency and Staging Change the Decision

Part-load performance is a system behavior. It includes compressor power, condenser fans or tower power, chilled-water pumps, condenser-water pumps where applicable, controls and the decision to bring another machine online. One low compressor kW number can be erased by poor staging or a low-ΔT condition that forces extra flow. EPA’s building-upgrade manual makes the same whole-system point: equipment, controls and operating practice interact, so a chiller-only metric isn’t a complete cost-effectiveness test.

DOE’s current chiller acquisition tables show why the test mode must be stated. For one positive-displacement water-cooled size band, the published values include 0.544 kW/ton at full load and 0.520 kW/ton for IPLV; the paired requirements exist because buyers can misjudge how a machine will actually run. Those values are procurement benchmarks, not a promise for your plant.

Put candidate curves on the same assumptions. A varying-load regime should show the behaviour between the design point and the minimum safe point:

1. Define the load bins and hours in each bin.
2. For each bin, list active compressors, commanded capacity and the pump or fan state entering the condenser.
3. Evaluate starters, dead times, minimum run times, minimum flow values and leaving-water-temperature stability.
4. If energy cost drives the project, calculate both chiller-only power and whole-plant power.

For a procurement review, treat a missing part-load curve as a documented risk. Request the manufacturer’s test evidence and map it to the operator’s sequence before a purchase order is released. In a real plant, the result depends on the controls engineer, the pump schedule and the factory settings—not on a single catalogue line.

A Pumps & Systems case history connects hydraulic changes with electricity cost and reports savings expected to approach 13.21% over a full year for that project. It is not a universal result, but it is a useful reminder that piping and control choices can change the value of a compressor selection.

That plant-side comparison matters: a chiller-only curve can look attractive while a pump or tower sequence erases the difference.

Air-Cooled or Water-Cooled? Keep Heat Rejection Separate

“Scroll or screw” and “air-cooled or water-cooled” answer different questions. The first describes compression; the second describes how the site rejects heat. Combining them into one label can hide important design assumptions.

Do not turn the result into a compressor-name contest. The ASHRAE liquid-chilling-system guidance treats entering conditions, flow and application inputs as part of the decision boundary.

Keep the decision boundaries visible

Boundary	Questions to answer	Evidence to request
Compressor architecture	How does capacity control behave from minimum to design load?	Capacity map, start/stop limits, sound and service notes
Heat rejection	Is condenser air or water available at the required conditions?	Ambient bins, tower approach, water quality, fan/pump curves
Hydraulics	Will flow and ΔT remain inside the control envelope?	Flow meter data, design ΔT, bypass and pump sequence
Site constraints	What do noise, water, footprint, plume, maintenance and codes allow?	Site survey, utility limits, acoustic and compliance review
Load profile	Where do the peak, typical and minimum loads sit?	At least a 12-hour trend with 5-minute load intervals
Rating basis	Are all options compared at the same water and condenser conditions?	Two operating points with included-power boundaries stated
Refrigerant	Which jurisdiction, safety and service assumptions apply?	Current compliance record and lifecycle availability check
Acceptance	How will the selected package be verified after startup?	Evidence: one 2-hour witnessed run and one 24-hour alarm review.

For a water-cooled option, include tower fans, condenser-water pumps, water treatment, makeup water and seasonal approach in the energy model. State the entering condenser temperature for every quoted rating and check the available hydraulic pressure, for example an 8 kPa allowance in a preliminary pump schedule. For an air-cooled option, include ambient design bins, coil fouling, fan control and the effect of hot outdoor air on lift. The choice may change the best compressor architecture because it changes the compressor’s condensing condition.

Put the compressor and heat-rejection choices side by side: an air-cooled scroll chiller and an air-cooled screw chiller both reject heat straight to outdoor air, so an air-cooled chiller, or air cooled chiller, of either compressor family shares the same fan and coil design constraints (a spec sheet may write it air cooled scroll or air cooled screw chiller without the hyphen, and reviewers read it the same way). A water-cooled scroll chiller and a water-cooled screw chiller instead send heat to a cooling tower loop. Whichever term a spec uses — water cooled screw chiller, one of several water-cooled screw chillers on a shortlist, or simply a water-cooled chiller — water cooled chillers and other water chillers on the market share that same tower and condenser-water dependency regardless of compressor type. In practice, most chiller systems on a given site are chosen air-cooled or water-cooled first, then scroll or screw second, because the heat-rejection path constrains the mechanical room before the compressor label ever matters.

The ASHRAE liquid-chilling-system material treats selection as a systems issue involving entering conditions, flow and other plant inputs.

How to Read COP, IPLV and kW/ton Without Being Misled

COP is a ratio of useful cooling effect to input power under a stated condition. kW/ton expresses input power per refrigeration ton; lower is better at the same boundary. IPLV is an integrated part-load metric built from weighted test points. These numbers are valuable, but none is an annual plant-energy guarantee.

Ask for entering and leaving chilled-water temperatures, entering condensing temperature or ambient, flow rate, fouling factors, voltage and frequency, included pumps or fans, and the edition of the standard. Confirm that the test temperatures fit your application. Glycol, hybrid or other fluids may require a different rating basis. AHRI 550/590 ratings answer a different question from a plant model; the HPAC central-plant example illustrates why one number can mislead when hydraulics and tower power are omitted.

Here’s a simple unit check. If a chiller draws 300 kW while delivering 1,000 refrigeration tons at the stated test point, its chiller-only intensity is 0.300 kW/ton. Using 1 refrigeration ton ≈ 3.517 kW of cooling, the corresponding COP is approximately 3.517 ÷ (0.300 × 3.517? No, keep the definitions straight): cooling output is 1,000 × 3.517 kW = 3,517 kW, so COP ≈ 3,517 ÷ 300 = 11.72. That conversion is mathematically correct for the stated boundary, but it doesn’t add tower, pump or distribution power.

When comparing two quotations, put both ratings in one table, normalize conditions and apply the same load bins. A 5% difference in entering conditions can change the result, so keep the source data rather than rounding it away. If a supplier cannot explain what is included in a number, mark the comparison as not yet comparable.

A facility manager can use a simple scenario: compare the same 6-hour operating window with and without condenser-water pump power. The measured evidence may reverse the apparent winner, which is why a factory test report and a site acceptance test should be read together.

The risk is a false efficiency comparison. Keep the calibrated test evidence, the factory report and the plant-model assumptions in the same review pack; otherwise a 1-hour data slice can be mistaken for an annual result.

Even a 5% change in entering conditions can move the result. A glycol or other process-fluid system needs its own rating basis, and the HPAC discussion shows why a single-number result can miss cooling-tower effects.

For clarity, retain the terms chiller-only, single-number and cooling-tower in the review notes so the included-power boundary cannot disappear during procurement.

Refrigerant and Compliance Questions to Put in the Spec

Refrigerant selection is a project constraint, not a marketing descriptor. Verify jurisdiction, safety category, charge position, leak-detection and ventilation rules, service-tool availability, technician training, recovery specifications and expected change-out life. “Low GWP” alone is not a complete compliance claim.

In the U.S., the EPA Technology Transitions table lists different dates and GWP thresholds for chiller subsectors. The obligation depends on equipment class, leaving-fluid temperature, installation date and charge, so do not copy a data-sheet statement from another product category into your specification.

The linked Technology Transitions sector table separates comfort-cooling and industrial-process cases; confirm which one applies before tender.

Put these fields into the procurement schedule:

• Refrigerant name or designation, safety class, and the applicable code or standard.
• Factory charge, field-charge boundary and leak-detection provisions.
• Required service equipment, recovery method and technician qualification.
• Jurisdiction-specific installation and phase-down dates, confirmed at tender stage and prior to shipment.
• What if refrigerant isn’t available over its estimated asset life?

The compliance risk is highest when a project team copies a rule from a different jurisdiction. Ask the local authority and the installation contractor to confirm the scenario, then retain the manufacturer’s factory charge record and the refrigerant safety documentation as evidence.

The 7-Point Load-Profile Decision Grid

Use this grid during a design review. Each point requests evidence and ends with a next action, so the decision does not depend on a show of hands or a familiar product label. For a first pass, review at least 12 hours of trend data and repeat the check after 24 hours. The evidence-first structure follows the measurement and commissioning logic in ASHRAE’s central chilled-water plant guidance.

7-Point Load-Profile Decision Grid

Point	Decision question	Evidence that closes it
1. Peak	What is the design-day load and how long does it last?	Heat-load calculation and duration assumption
2. Typical band	Where do most operating hours sit?	Hourly bins or an auditable proxy
3. Minimum	What is the lowest stable load and flow?	Trend data, minimum-flow limits and cycling log
4. Step changes	Are there batch or recovery events?	Process sequence, buffer volume and recovery target
5. Heat rejection	What condenser condition is available in each season?	Ambient/tower bins, water quality and utility limits
6. Controls	Which machine starts, unloads or stops—and why?	Sequence of operations and trend points
7. Proof	How will performance be accepted after startup?	Witness test, data export, alarms and sign-off owner

If a site is missing an answer, the next step is a bounded test—not “screw wins” or “scroll wins.” Capture flow for 2 weeks, obtain a part-load map, verify the tower approach or revisit the design-day assumption. Without a full season, a clearly labelled sample month is still more useful than an unqualified extrapolation. That is the purpose of the grid: a decision framework, not a product list.

Commissioning: Build a Commissioning Evidence Trail

Commissioning should leave an operating record that remains useful after the project team departs. A start-up sheet without calibrated, traceable measurements of flow, entering and leaving temperatures, pressure drop, electrical input, condenser condition and operating state records attendance, not performance. Set a review window—often within 24 hours of first start and again after a representative 2-hour run—so exceptions are not buried in closeout files. An owner may also set a 1-hour baseline, a 2-month seasonal review and a 4% alert threshold based on operating risk.

Use a Commissioning Evidence Trail with five layers:

1. Identify: verify serial number, refrigerant, firmware, sensors and options vs approved submittal.
2. Hydraulics: record entering/leaving temperatures, flow, pressure drop, pump condition and strainer condition.
3. Heat rejection: record outdoor or condenser-water condition, fan or tower condition, approach and alarms.
4. Controls: trend load command, compressor condition, capacity signal, setpoint, staging decision and fault history.
5. Follow-up: repeat selected points under a representative load, record any 30%-or-greater change in commanded capacity, and attach the export, exceptions and owner.

ASHRAE central chilled-water plant material and its commissioning guidance focus on functional testing and documented turnover. The objective isn’t to impose a uniform template on each site; it’s to make the acceptance boundary transparent.

Independent field comparisons support the caution. The GSA/ORNL variable-speed screw evaluation found similar results to its comparison technology and concluded that site-specific conditions determine the most cost-effective selection. That is a useful balance to any “screw always wins” or “scroll always wins” headline.

Commissioning evidence type matrix

Evidence type	Minimum capture	Owner and timing
Identity	1 serial-number and refrigerant check	Commissioning lead before energization
Flow	2 stable readings at design and minimum flow	Controls technician during functional test
Temperature	Entering/leaving values at 5-minute intervals	TAB engineer on witness day
Heat rejection	3 ambient or condenser-water conditions	Mechanical contractor in trend log
Controls	Sequence, setpoint and capacity command	BMS owner at 15-minute trend interval
Alarms	Active, cleared and acknowledged alarm list	Operator at turnover
Repeat test	2-week representative-load follow-up	Owner’s engineer after occupancy or production
Closeout	8-point exception and sign-off record	Project manager at final handover

What the 2026 Buying Conversation Should Ask

Exact-keyword history for “Scroll / screw chiller” is too limited for a quantified market forecast in this run. The current 2026 signals used here are integration and compliance planning: clients want to know how equipment interfaces with controls, utilities, refrigerant rules and lifecycle evidence. Treat that as a procurement question, not a market-size claim.

Chiller requirements differ by facility type: commercial buildings often prize quiet operation and steady comfort cooling, data centers and other industrial applications typically need continuous, higher cooling demand and fewer moving parts to reduce failure points, and a plant chasing larger cooling capacity growth may need staged circuits rather than one oversized unit. These differences in chiller design, not a universal ranking of scroll chillers and screw chillers, decide the right chiller for a given site. Energy savings and energy consumption goals point the same way: matching the chiller unit to the real cooling needs of the building, not to a nameplate, is what an industrial chiller systems review should confirm before anyone leans on a screw chiller models comparison chart to choose a scroll or screw unit alone.

Ask every shortlisted supplier:

• Could you supply the part load map and sequence, not only the design point?
• What power is included in each rating—compressor, fans, pumps or a complete plant?
• What data points are exposed to the site BMS, and at what interval?
• Which refrigerant rules and service assumptions were used, and where must the local authority affirm them?
• What will be in the handover package if the operating load is not equivalent to the design estimate?

For first-party context only: Koven Air’s public company information states a screw-chiller milestone in 2007, own-brand export from 2010, a 20,000 m² facility, traceability and in-house testing. These details explain why test evidence is a useful conversation starter; they are not standardized performance claims. Read the company history separately.

For the company background, see About Koven Air; the page is context, not a substitute for the site-specific part-load or commissioning evidence.

When to Hand the Decision to the Commercial Page

After the load grid, rating boundary, refrigerant scope and commissioning specifications are determined, transition to the project-specific commercial discussion. Koven Air’s scroll and screw chiller solutions page is the suitable place to evaluate product variation and request a scoped response. This report should continue to be the objective information layer; it shouldn’t convert the load grid into a disguised catalog.

If you require a project discussion, use the site’s contact prompt to open the Koven Air enquiry window. Present the seven load-grid points, not only a preferred compressor designation.

At this handoff, the buyer should attach the load-profile evidence, the rating boundary and the commissioning acceptance scenario. Koven Air’s factory and engineering information can then be checked against those facts, while the commercial page handles the product-specific response. Keeping those two layers separate reduces comparison risk and preserves a clear audit trail.

Frequently Asked Questions

What is the difference between a scroll chiller and a screw chiller?

A scroll chiller uses one or more orbiting-scroll compressors, while a screw chiller uses intermeshing-rotor compressors. Both are positive-displacement vapor-compression machines. The difference is therefore a compressor mechanism, not a complete plant specification. Capacity control, number of circuits, staging, condenser type, refrigerant, pumps, fans and controls can change the result just as much as the compressor family. Compare the machines against the same load bins, temperatures, flow and included-power boundary before choosing.

What is a scroll chiller?

In a scroll chiller, refrigerant compression is performed by scroll compressors. The orbiting scroll traps refrigerant vapor in pockets and compresses it toward the discharge port. Scroll packages are often arranged with multiple compressors or circuits so the controls can stage capacity. That arrangement can be useful for changing loads, but it still needs a verified minimum-load strategy, start/stop logic, flow protection and a clear sequence of operations. “Scroll” alone doesn’t tell you the expected annual energy use.

How does a screw chiller work?

With a screw chiller, two intermeshing rotors move and compress refrigerant vapor. Capacity may be controlled with a slide valve, variable speed, stepped circuits or a combination. The machine rejects heat through its condenser and returns cooled water through the evaporator, so the compressor is only one part of the operating boundary. Ask for the compressor map, oil-management requirements, minimum stable load, entering-condenser assumptions and the control sequence that determines when the screw unloads or another machine starts.

Which chiller type is most efficient?

There’s no architecture that’s most efficient at every load, climate and plant boundary. One rating can favor an option at a stated test point while the annual plant result favors another after pumps, fans, tower approach, cycling and controls are included. Normalize the conditions, model the load bins and ask for both full-load and part-load evidence. If the decision is based on energy cost, compare whole-plant power and the expected operating hours, not just one COP or IPLV figure. Also ask whether the quoted number includes the condenser fan, tower, pumps and controls. One small difference in chiller-only COP may disappear when a tower operates at a different approach or when low ΔT forces extra flow. Efficiency is therefore a measured outcome of the selected plant and sequence, not a permanent attribute of one compressor word.

What are the four main parts of a chiller?

The vapor-compression core is commonly explained as compressor, condenser, expansion device and evaporator. The working industrial package also includes controls, sensors, water-side heat exchangers, pumps or fans, safety devices, piping, electrical equipment and a heat-rejection system. The four-part explanation is useful for learning the cycle; it isn’t enough for procurement or commissioning. The complete specification should state the water temperatures, flow, condenser conditions, refrigerant, control mode and acceptance evidence around that core. On a real handover, confirm that the sensor list, alarm list and trend interval are attached; a component diagram without those records can’t prove stable operation.

What are common scroll and screw chiller selection mistakes?

The recurring mistakes are procedural: selecting from the design peak only, treating a compressor rating as a plant rating, ignoring minimum stable load, running too many machines at low load, and leaving heat rejection or refrigerant compliance until late in the project. Another mistake is asking for “the most efficient chiller” without defining the test boundary. Use the seven-point grid, record the unknowns and require a commissioning trail.

References & Sources

The following documents supported the definitions, rating caveats, compliance examples and operating-method guidelines. They’re hyperlinked for confirmation; Koven Air’s commercial documents are listed separately as internal handover targets.

• AHRI 550/590 and 551/591 performance-rating scope.
• U.S. Department of Energy FEMP: Purchasing Energy-Efficient Electric Chillers.
• ASHRAE Handbook: Liquid-Chilling Systems.
• ASHRAE: Design and Control of Central Chilled-Water Plants.
• Pumps & Systems: Controlling Electricity Costs and Temperature Exchange.
• HPAC Engineering: Central-Chiller-Plant Modeling.
• U.S. EPA: Technology Transitions HFC Restrictions by Sector.
• U.S. EPA ENERGY STAR Building Upgrade Manual.
• GSA/ORNL Variable-Speed Direct-Drive Screw Chiller evaluation.
• Google Patents: US20240102472A1 refrigerant screw-compressor improvement record.

Inter-service handoff: Koven Air scroll and screw chiller solutions About Koven Air