Beverage filling line OEE is useful only when the plant defines what the line includes, when production is supposed to run, which speed applies to each SKU and where good output is counted. Without those rules, two shifts can report different OEE figures for the same physical performance. A consistent method turns the percentage into evidence for production and equipment decisions.

What OEE measures on a beverage filling line
Overall equipment effectiveness combines three factors:
OEE = Availability × Performance × Quality
- Availability measures how much of the planned production time the line was running.
- Performance compares actual output during run time with the output possible at the qualified ideal rate.
- Quality measures first-pass good output as a share of total output.
The standard calculation described by OEE.com expresses the components as:
Availability = Run Time ÷ Planned Production Time
Performance = (Ideal Cycle Time × Total Count) ÷ Run Time
Quality = Good Count ÷ Total Count
The same OEE result can be expressed as fully productive time divided by planned production time. Keeping the three factors separate is more useful because it shows whether the primary loss comes from stops, reduced speed or rejected output.
OEE is a loss-measurement tool, not a direct statement of profitability or total plant capacity. Product demand, labor, utility limits, sanitation schedules and unused calendar time require separate measures. Define the question first: a team improving one filler needs a different boundary from a buyer evaluating the output of a complete beverage production line.
Set the measurement boundary before collecting data
Draw the boundary from the first process or packaging step included in the analysis to the agreed good-product counting point. On one project, that may cover the rinser-filler-capper. On another, it may include bottle making, filling, labeling, secondary packing and palletizing.

Choose a line output counter after the last operation that determines whether the unit is saleable within the boundary. A bottle that leaves the filler correctly but is later rejected for a missing cap or unreadable label should not appear as good line output when capping and labeling are included.
Also identify the line constraint for each product and package format. The constraint may move when bottle size, label, pack pattern or beverage changes. A filler can be running efficiently while the complete line loses output because a downstream machine repeatedly blocks it. Research on line-level effectiveness in beverage bottling similarly distinguishes a production-line measure from the OEE of one machine.
Record the boundary, count location, product, package, shift and time basis beside every OEE result. That information makes comparisons traceable.
Calculate availability, performance and quality consistently
The calculation is simple. The definitions behind each input need more care.
| Factor | Required data | Beverage-line losses captured |
|---|---|---|
| Availability | Planned production time and stop time | Breakdowns, changeovers, sanitation, material shortages and other stops inside the defined production window |
| Performance | Run time, total count and qualified ideal cycle time | Reduced speed, short stops, repeated blocking and starvation that do not cross the stop threshold |
| Quality | Total count and first-pass good count | Fill, closure, container, coding, label and pack rejects inside the boundary |
Planned production time starts after schedule losses have been removed. If a shift was never scheduled to produce, that time is outside OEE. When a changeover or cleaning activity occurs during time assigned to production, classify it consistently as stop time. Moving the same event in and out of the calculation makes the trend unreliable.
Set a stop threshold before data collection. Events at or above that threshold enter the availability record with a reason and duration. Shorter events remain visible through performance loss. The threshold can follow the plant’s reporting resolution, but it must remain stable across the periods being compared.
Choose the correct ideal rate for each SKU
Performance needs an ideal cycle time or ideal rate for the actual product and package being run. A budget rate, recent average or speed reduced to protect a weak downstream machine will hide part of the loss. The selected ideal rate should represent the fastest qualified rate for the defined line and SKU under correct operating conditions.
For a shift with several SKUs, calculate performance with the ideal cycle time for each run:
Performance = Σ(Ideal Cycle Timeᵢ × Total Countᵢ) ÷ Total Run Time
The OEE multi-product guidance recommends retaining the underlying run time, planned time, ideal cycle time, total count and good count for each product. A simple average of SKU-level OEE scores can distort the result because the runs may have different durations.
Create a controlled rate table by product, container, closure and pack format. Store the revision and qualification basis with each rate. If performance exceeds 100 percent, the ideal cycle time, count signal or time basis is wrong.
Classify line stops without double counting
A line can stop because a machine faults, because it has no incoming containers or because its discharge is blocked. The event log should distinguish these states. Otherwise, the same incident may be charged to the filler, labeler and packer even though one root event stopped the complete line.

Use one primary line-stop record with start time, end time, affected SKU, observed location and reason. A cause hierarchy can then connect the line stop to the responsible machine or external condition. Useful top-level groups include equipment fault, changeover, sanitation, lack of material, quality hold, utility interruption and planned test.
Blocked and starved states deserve their own fields. They show how one machine’s behavior propagates through conveyors and accumulation. A filler that is blocked by a packer should not receive the same diagnostic label as a filler with an internal fault.
Reason codes should be short enough for operators to use and detailed enough for action. An excessive list produces uncategorized entries; a list containing only “machine stop” gives maintenance and production teams nothing to investigate.
Define quality as first-pass saleable output
OEE quality uses first-pass good count. Units that need rework are not first-pass good output, even if they are recovered later. Track recovered product separately so OEE remains comparable while the plant still accounts for final yield.
Define reject rules for every operation inside the measurement boundary. Depending on the line, these may include damaged containers, incorrect fill, closure defects, coding failures, label faults and failed secondary packs. The good counter and reject counters should reconcile with the total count over the same time window.
Delayed laboratory results need a closing rule. The plant can hold the shift result open until quality data arrives or publish a provisional result that is revised later. The method and revision status should be visible in the report.
Build a shift-level data table
A manual study can begin with a spreadsheet or controlled shift sheet. Automated collection improves resolution, but automation does not repair unclear definitions. Start with these fields:
| Field | Purpose |
|---|---|
| Shift and production order | Links the record to the schedule |
| Product and package SKU | Selects the correct ideal rate and reject rules |
| Planned start and finish | Establishes planned production time |
| Run and stop timestamps | Calculates availability and event duration |
| Line state and reason code | Separates faults, blocked, starved, changeover and sanitation states |
| Source machine | Locates the initiating event without duplicating line loss |
| Total count and good count | Calculates performance and quality |
| Reject reason | Directs quality investigation |
| Ideal cycle time revision | Preserves the calculation basis |
| Operator comment | Adds context for unusual events |
Review missing time, overlapping events, counter resets and unexplained performance above 100 percent before releasing the result. A daily loss review should focus on the largest verified loss category, not on the headline percentage alone.
Use OEE to guide line and equipment decisions
The three factors lead to different actions. Availability loss may point to fault elimination, faster changeovers or better material coordination. Performance loss directs attention to short stops, unstable transfers, speed settings and the actual constraint. Quality loss calls for defect analysis at the process step that creates the reject.

Use the loss tree before proposing new equipment. A faster filler will not raise line output when the current constraint is the labeler, packer, utility supply or frequent format change. Compare alternative projects by their effect on the verified constraint and loss category.
A single universal OEE target is a weak purchasing specification. NIST material referencing ISO 22400 notes that OEE combines availability, effectiveness and quality, and that benchmark comparisons require comparable work units and processes. Establish a baseline and improvement target for the same product mix, boundary and calculation rules. External figures can provide context but should not replace plant evidence.
Specify OEE data during an equipment project
Data requirements belong in the request for quotation, control specification and acceptance plan. Add the following items to the project documents:
- machine states and the conditions that trigger them;
- count signals, counter locations and reset behavior;
- timestamp resolution and synchronization across machines;
- fault and stop-reason structure;
- product recipes and the approved ideal rate for each format;
- reject signals and the definition of first-pass good output;
- data access, export format, retention and ownership; and
- the test method used to confirm signals during FAT and SAT.
A common machine-state model can simplify a line built from several equipment packages. OMAC describes PackML as a standard that supports consistent machine behavior and data retrieval across different control systems. Buyers should specify the required standard or interface explicitly; this article does not assume it is included in any machine.
Use Sheenstar’s quotation checklist to define product and package scope, then carry the approved data points into the site-preparation checklist and FAT checklist. This creates one measurement basis from quotation through commissioning.
Questions to settle with the line supplier
Before approving the control and reporting scope, ask:
- Where are total output, good output and rejects counted?
- Which signal defines line running, stopped, blocked and starved?
- How are simultaneous machine faults assigned to one line-stop event?
- Which ideal rate applies to every agreed product and package?
- Can operators add or correct reason codes without overwriting the event history?
- Which data can the plant export for independent analysis?
Sheenstar’s liquid filling line guide and can filling machine buying guide provide equipment context for different line types. For a project-specific discussion, send the beverage, package formats, target output, proposed line boundary and data requirements through the contact page.
Frequently asked questions
What is a good OEE for a beverage filling line?
There is no universal figure that makes dissimilar lines comparable. Product mix, sanitation, changeovers, package formats and the measurement boundary all affect the result. Use a stable internal definition, compare like-for-like periods and set targets from verified losses on the same line.
Should cleaning and changeover time be included in OEE?
Include an event as availability loss when it occurs inside time assigned to production and the line was expected to run under the plant’s definition. Time that was never scheduled for production is normally a schedule loss outside planned production time. Apply the rule consistently.
Is filler OEE the same as complete-line OEE?
No. Filler OEE describes the selected filler boundary. Complete-line OEE follows output through the agreed line boundary and includes the effect of upstream starvation, downstream blocking and rejects created after filling.
How is OEE calculated when several SKUs run in one shift?
Retain an ideal cycle time and counts for each SKU. Sum the ideal production time for all runs, then divide by total run time for performance. Calculate availability and quality from the corresponding combined planned time and counts.
Can OEE exceed 100 percent?
No. A value above 100 percent indicates an incorrect ideal rate, count, time base or data mapping. Investigate the calculation instead of capping the displayed value.
Which OEE data should be checked during FAT?
Check machine and line states, count signals, reject signals, timestamps, reason codes, recipe-specific ideal rates, data export and counter reset behavior. The FAT method should state the test conditions and expected evidence for each signal.