How to Balance Printing, Bag Making, Handle Attaching, and Packing Capacity

A non-woven bag line does not have one capacity. It has a chain of capacities that change with bag design, order mix, quality holds, staffing, material availability, and the way work is released. Quoting the fastest station’s nameplate speed as the line capacity is an easy way to create missed delivery promises, excess work-in-process, and arguments between departments.

This guide explains a practical method for planning and improving a PP non-woven bag operation that includes printing, converting, handle attaching, inspection, and packing. It is for export buyers evaluating a supplier and for factory teams trying to make more reliable promises. The calculations are illustrative. They must be rebuilt from the actual product routing, approved specification, machine configuration, safety limits, staffing model, and historical data.

Capacity begins with the customer order, not the machine brochure

A buyer’s order contains a quantity, delivery date, bag dimensions, fabric construction, printing requirement, handle type, packaging configuration, inspection requirement, and sometimes a staged shipping plan. Each feature changes work content. A plain carry bag routed directly to converting and packing cannot be planned like a multi-color printed bag with a separate handle attachment and individual insert. Even two bags with identical dimensions can have different constraints because one requires careful print registration and the other has a reinforcing patch that adds handling time.

Translate the order into a routing card. List each process step, required machine or work cell, changeover class, standard approved rate range, expected yield, inspection points, buffer rule, pack-out method, and whether the step can run in parallel. Do not hide manual work under “packing.” Counting, folding, inserting, carton preparation, sealing, labelling, palletizing, and rework may each consume scarce people or space. If a buyer expects a barcode label, special master carton, or pre-shipment inspection sample, include it in the routing instead of treating it as an afterthought.

The objective is a realistic promise interval, not a heroic daily maximum. A capacity plan should state its assumptions: operating hours, planned breaks, approved speed band, expected availability, product mix, yield, staffing, changeovers, material readiness, and contingency. If an assumption changes, the promise should be recalculated. This is far more credible than saying a line can produce a single large number without defining which bag, which shift pattern, or which quality standard.

Find the constraint by observing flow, not by voting

The constraint is the resource or condition that limits the whole system’s output for the relevant product mix. It is not always the slowest-looking machine. A fast bag-making machine can be constrained by a handle cell that accumulates unprocessed bags. A handle cell can be constrained by printed web availability. A packing station can become the constraint when an export order requires small retail bundles, strict carton counts, or labor-intensive labelling. Material shortages, inspection releases, and frequent product changeovers can also be the effective constraint.

Walk the order through the factory. Observe where work waits, where people rush to recover, where stockouts occur, which station runs during overtime, and which station starves despite downstream queues. Use a time-stamped count of good units entering and leaving each step. Measure for more than one hour and more than one product, because an isolated observation can confuse a temporary break or material change with a lasting constraint. The data do not need to be sophisticated; they need common definitions and sequence.

Use good output, not gross output, in every comparison

Gross speed is seductive because it is easy to display. The buyer, however, receives good bags packed to the approved requirement. Build every capacity calculation around good output: units that have completed the required route, passed applicable checks, and are available for the next approved step. Record scrap, rework, holds, and first-pass yield separately. A station that produces a high gross count while creating a quality queue is not increasing shipment capacity.

ISO 22400 provides a framework for defining and using manufacturing operations KPIs, including formulas and characteristics for selected measures. ISO’s description explains that the standard addresses KPIs for manufacturing operations management. Use this as a discipline rather than as a certification claim: define each KPI, source data, unit, time boundary, exclusions, owner, and decision it will support. Do not compare a printed-web meter count with a finished-bag count without converting them to a common product and quality basis.

A simple daily board can display planned good output, actual good output, current constraint, oldest WIP age, quality holds, and critical material status. The board should explain what operators can do with the information. For example, if packing is the constraint, accelerating upstream converting may only increase inventory and damage risk. The correct response may be to protect the packing team with cartons, approved labels, staffing, and a controlled upstream buffer.

Calculate available production time honestly

Begin with scheduled time, then subtract planned non-production time such as breaks, planned meetings, scheduled maintenance, known changeovers, and product-specific cleaning where these are part of the standard route. The remaining planned production time is not automatically productive time. Unplanned stops, waiting for material, quality checks, staffing gaps, and speed loss reduce it further. Avoid a single “efficiency factor” that conceals the causes. Keep availability, rate, and quality loss visible enough to act on.

For an illustrative example, assume a handle station has 420 scheduled minutes in a shift. It has 40 minutes of planned changeover and inspection preparation, leaving 380 planned running minutes. If 35 minutes are lost to documented interruptions, 345 minutes remain. If the approved actual good rate for the bag family is 48 bags per minute, its theoretical good-output opportunity for that condition is 16,560 bags before considering additional yield loss. This example is not a promise for any machine. It is a transparent calculation whose inputs can be challenged and improved.

Repeat the calculation by product family. A rate observed on a simple flat bag should not be used to sell a complex box bag with online or offline handle work. Use the approved sample and routing to decide what constitutes one complete bag. If material is ordered in kilograms or meters, include an approved conversion and expected scrap factor; never quietly convert a material quantity into sellable bags without reconciling yield.

Separate technical rate from operational capacity

Technical rate is what a station can achieve in a controlled condition. Operational capacity is what the whole factory can reliably deliver during normal order execution. Operational capacity includes schedule adherence, changeovers, product release, materials, maintenance, operators, quality holds, and packing. Both numbers are useful, but they answer different questions. The first informs equipment selection and process capability. The second informs customer delivery commitment.

Make a rate card for each product family. Include bag style, fabric range, print complexity, handle construction, number of operations, target speed range, common defects, setup time, staffing, quality checks, and approved packing route. Mark the evidence date and conditions. When a new order falls outside the card’s range, label the estimate as provisional and plan a sample or trial. This protects both the supplier and buyer from treating an old production rate as a universal guarantee.

Plan buffers as protection, not storage

Buffers are useful when they protect the constraint from predictable short interruptions. They are harmful when they hide defects, mixed lots, or a scheduling error. Define a buffer in good units, physical location, maximum age, lot-segregation rule, owner, and replenishment signal. Place it where it protects the constraint without creating excessive handling. If printing is intermittently unavailable because of long setup changes, a controlled printed-web buffer may protect converting. If handle attachment is the constraint, an upstream bag buffer may be appropriate only if bags remain identifiable, clean, and free from crush or mix-up risk.

Do not create a buffer for a quality hold. Held product needs a clearly marked, access-controlled location and disposition process. Mixing held and released bags to keep a station busy creates a larger commercial risk. Likewise, do not push work into packing merely to make upstream utilization look good when cartons, labels, or release evidence are not ready. A flow system protects the customer’s order, not an individual department’s utilization score.

Use product families to control changeover loss

Changeovers consume more than machine time. They can require roll changes, ink or print setup, tool change, recipe selection, first-piece approval, cleaning, carton substitution, label verification, and operator briefing. Group orders into sensible product families where this reduces changeover without violating due dates, material segregation, or buyer commitments. Document which features drive the change: fabric color, print artwork, bag dimensions, handle tool, packaging, or inspection requirement.

Measure changeover from the last good conforming unit of the prior job to the first approved conforming unit of the next job. This definition exposes waiting time and first-piece approval, which are often excluded from a machine-only stopwatch. Break the activity into external work that can be prepared while the line runs and internal work that requires the line to be stopped. Pre-stage approved fabric, artwork files, tooling, cartons, labels, and quality samples when safe and practical. Do not pre-stage items in a way that creates material mix-up risk.

Balance people, machines, and quality decisions together

Line balancing is not only a machine problem. A semi-automatic operation can have the correct machine speed but insufficient people for collection, alignment, handle placement, visual inspection, packing, and material replenishment. Conversely, adding people to a non-constraint station can increase WIP rather than shipment output. Map manual elements in seconds per bag and in batch tasks per carton or pallet. Include walking distance, reaching, reorientation, inspection, label verification, and rework handling. Small repeated motions become significant at export-order volume.

Quality staffing needs the same realism. If a product requires a first-piece approval at every change, in-process checks, and final packing review, make those activities available in the schedule. Pressure to skip checks is often a sign that the planned capacity assumed quality work did not exist. The correct remedy is to redesign the schedule, method, or resource—not to redefine quality as downtime after the order is late.

Promise dates using a finite-capacity conversation

When sales receives an enquiry, it should provide production with the actual customer requirements, not only a quantity. Production then checks approved rate-card coverage, material lead time, current committed load, constraint availability, required trials, changeovers, inspection, packing, and shipping cut-off. The answer can be a firm date, a conditional date subject to sample approval, or a request for a staged shipment. A conditional answer is not weak if it identifies the true decision point early.

Communicate the basis to the buyer. For a Customizable or OEM order, explain that capacity will be confirmed after the approved sample, routing, and material plan are fixed. Do not use a fast demonstration run to imply that every future order will ship at that pace. For repeat products, use historical good-output performance and current finite load, then retain a modest contingency consistent with the buyer agreement. This approach is more likely to improve on-time delivery than optimistic quoting followed by overtime and rushed inspection.

Case study direction: test whether protecting the real constraint improves delivery reliability

This is a study design, not a reported Zhengxin customer case. Select a defined bag family for which late delivery or excess WIP is a recurring issue. For a baseline period, collect order lead time, planned and actual good output at each route step, WIP age, changeover time, holds, constraint interruptions, overtime, and on-time shipment. Identify the apparent constraint from observed flow rather than assumptions. Then introduce one limited intervention: for example, a controlled upstream buffer, pre-staged changeover kit, clearer release rule, or constraint-focused maintenance window.

Compare periods only after controlling for major product mix changes. The primary outcome may be on-time shipment or stable good output at the constraint; secondary outcomes can include WIP age, expedited movements, scrap, and overtime. Report trade-offs honestly. A buffer may reduce starvation but increase handling. A setup kit may reduce changeover but require more disciplined material verification. The study is useful if it reveals which mechanism changed, not merely if a headline KPI improves for one week.

Review the plan every week, but avoid daily panic rescheduling

A weekly capacity review should compare demand by ship date with finite capacity by product family, constraint, material status, and packaging readiness. Flag orders that need a buyer decision, sample approval, alternate material approval, extra shift, or shipment split. Update the plan when a verified event changes it: a machine failure, material delay, quality hold, or confirmed rate change. Do not rebuild the schedule every time a local station experiences a short interruption; that simply transfers noise through the factory.

Use a daily tier meeting for actual execution: what was the previous day’s constraint, what is today’s critical order, what material or quality release is blocking it, and what decision owner will remove the blocker? Keep the meeting evidence-based and short. The factory does not need a complex digital twin to start. It needs one shared version of the order status and the discipline to measure good output where it matters.

Model order mix before accepting the average rate

Average rate is dangerous when the mix is lumpy. Suppose a factory makes three families: a simple unprinted carry bag, a printed bag with registration requirements, and a structured bag with a more complex handle operation. A blended monthly average can make capacity look comfortable even when the next two weeks contain an unusually high proportion of the structured family. Build the plan from product-family minutes at the constraint, not from total bags alone. Each order consumes constraint time according to its own approved route and observed good rate.

For each family, calculate estimated constraint minutes as planned good quantity divided by a validated good-output rate, then add a proportionate setup and risk allowance. Sum the demand by week and compare it with finite available constraint minutes. The excess is not an accounting problem; it is an early decision signal. Options might include moving a non-critical order, splitting a shipment with buyer agreement, adding a qualified shift, approving an alternate route after validation, or declining an unrealistic date. The correct action depends on commercial commitments and safe operating limits.

Keep estimates separate from measured results. A quotation estimate may use the closest approved family and a conservative rate range. Once a trial is approved, replace the estimate with observed data and update the capacity plan. This prevents an initial sales assumption from becoming an unchallenged operating fact. It also gives the buyer a fair explanation if a new construction requires a longer lead time than a visually similar prior bag.

Use a loss tree to make improvement choices

When output is below plan, ask where good units were lost. A simple loss tree starts with scheduled time and branches into planned downtime, availability loss, speed loss, quality loss, and non-flow loss such as waiting for material, release, or packing. Each branch needs a mutually understood definition. A line stopped for an approved artwork decision is not the same problem as a mechanical breakdown; both reduce delivery capacity, but their countermeasures belong to different owners.

For printing, common categories may include artwork approval, ink preparation, registration adjustment, cleaning, roll change, print defects, and drying or handling limits. For bag making, categories may include feed instability, cut or seal quality, tool change, web tracking, and material splice. For handle attachment, categories may include component replenishment, alignment, attachment quality, and operator motion. For packing, categories may include carton shortage, label mismatch, count verification, inspection hold, and pallet space. The categories should reflect the factory’s actual process, not a generic consulting list.

Prioritize a loss only when its frequency, duration, and influence on the constraint justify attention. A rare long incident may need reliability work; a short repeated interruption may need a simple replenishment signal or ergonomic change. Test countermeasures with a clear before-and-after definition. Avoid declaring an improvement after moving the loss to an upstream or downstream department. The true test is whether the order’s good output, lead time, quality, and safety improved together.

Connect material planning to line balance

Capacity disappears when the right fabric, printed web, handle component, carton, or label is not at the point of use. Material planning should therefore use the same routing and lot rules as production planning. Confirm incoming lead time, inspection status, approved substitute rules, roll width, consumption factor, expected scrap, minimum buffer, and storage conditions. For printed orders, confirm artwork release and color/registration approval before reserving critical line time. For special cartons or labels, treat proof approval and delivery as capacity prerequisites rather than warehouse details.

Use a readiness checklist before releasing a job to the constraint. It may include released work order, approved sample, material lots available, current recipe and tool, packing material, labels, quality plan, and staffing. A job that fails the checklist should not occupy the constraint merely because its due date is close. It may be better to run a ready compatible job and escalate the missing decision than to create a half-finished, hard-to-trace queue.

Make handoffs visible between processes

Each handoff needs a definition of done. Printing may release web only after a documented visual and registration check. Bag making may release output only after the specified dimensions and seams are accepted. Handle attachment may release only after applicable attachment checks. Packing may release only after count, carton, label, and lot identity are verified. A handoff standard should be concise enough for production use and strict enough that a downstream team does not discover a problem after hours of additional work.

Where products travel in tote bins, racks, or carts, label the container with product, finished lot or run link, quantity, status, and destination. A container should never carry both accepted and held product. Limit the number of containers waiting at each point. A physical visual limit makes overload visible; an unlimited floor area turns late problems into invisible inventory. If the line must deliberately build ahead for a later handle or packing window, document the reason and the planned consumption date.

Protect safe operating limits while pursuing flow

Pressure to meet an export cutoff can tempt teams to bypass inspection, extend maintenance beyond safe limits, add untrained labor, or alter guarded equipment quickly. These actions can create a more serious loss than the late order. Capacity improvement must remain within the equipment manual, workplace risk controls, qualified staffing, and approved quality method. A fast run is not a successful run if it produces untraceable, unsafe, or nonconforming output.

Build planned maintenance and training into the capacity calendar. A short protected maintenance window at the constraint may be more valuable than an unplanned failure during a critical order. Similarly, cross-training should include verified competence and defined task authorization, not merely placing another person at the station. The factory’s most credible promise is its ability to sustain a safe, controlled process, not to temporarily exceed it.

Turn daily learning into future quote accuracy

At order completion, compare quoted lead time and capacity assumptions with actual performance. Record the product family, route, actual good rate, changeover, material issues, quality loss, packing effort, and any buyer-driven changes. Update the rate card only after review; do not let one unusually good or bad run overwrite the baseline. Over several orders, this creates a more reliable database for sales and production planning.

Share the learning in a short cross-functional review. Ask which assumption was wrong, which information was missing at order entry, where the constraint moved, and which action would help the next similar order. This closes the loop between factory evidence and commercial promises. It also gives an international buyer a stronger reason to trust a supplier’s delivery plan: the plan is based on controlled learning rather than generic speed claims.

Before revising a committed date, identify the exact reason and its evidence. A material delay should be supported by supplier confirmation and revised arrival timing. A capacity shortfall should identify the affected product family and constraint minutes. A quality hold should identify the lot boundary and release condition. This precision makes customer communication more useful: it turns an excuse into a decision request, such as approving a split shipment, changing a pack-out date, or waiting for a validated recovery plan.

Use scenario planning for the largest orders. Build a base plan, a credible downside plan for one significant interruption, and an escalation plan that stays within qualified staffing, safety rules, and agreed quality controls. Name the trigger that moves the order from one plan to another. The factory then avoids the common pattern of discovering risk only after all available time has been consumed.

Frequently Asked Questions

What is the best capacity number to show an overseas buyer?

Show a product-specific, condition-based operational capacity range or delivery plan rather than a universal nameplate speed. State the bag family, route, shift pattern, quality basis, packing assumptions, and whether the figure is confirmed by an approved sample or historical repeat production.

Is the station with the lowest speed always the bottleneck?

No. The effective constraint is the resource or condition limiting flow for the current mix. It can be a handle cell, printing setup, packing labor, material availability, quality release, or a repeated changeover. Observe good-output flow and WIP before deciding.

How much buffer should be kept between bag making and handle attaching?

Use only enough identified, protected good work to shield the constraint from predictable variation. Define the maximum quantity, age, location, lot rule, and replenishment signal. Do not use a buffer to hide quality holds or unlimited overproduction.

Can OEE alone determine delivery capacity?

No. OEE-style measures can help understand a work center, but delivery capacity also depends on routing, product mix, changeovers, materials, labor, inspection, packing, and the constraint across the full order flow. Use common, documented KPI definitions.

How should custom bag orders be quoted before a sample is approved?

Quote capacity and delivery conditionally. Identify which factors need confirmation—material, artwork, handle construction, route, speed range, packing, and inspection—and convert the approved sample result into a controlled rate card before making a firm volume promise.