How to Estimate Cup Circulation, Return Points, Washing, and Loss Rates Before Launching a Borrow-and-Return Cup Program

A campus coffee shop planning a borrow-and-return cup program should model four numbers before buying a single cup: how many reusable cups must be in circulation (daily borrow-cup demand × average days out of circulation × safety buffer), how dense the return-point network must be, how much washing capacity the peak return flow requires, and what share of cups will leave the system permanently. Starbucks' Seattle tests illustrate each variable: its 2021 five-store Borrow A Cup pilot used a $1 refundable deposit, return kiosks, and professional washing that returned cups to circulation within 48 hours; its SODO 8 headquarters café has served every order in returnable cups by default since early 2022. The 2024 citywide Petaluma program — 30 businesses, 60+ return bins, a 51% return rate — shows what happens when the loop is open rather than closed. This article turns those cases into working arithmetic — sizing formulas, placement checklists, wash-capacity math, loss tracking, a four-week pilot with acceptance thresholds — and ends with the hybrid conclusion most campus operators reach: a reusable loop covers stay-on-campus regulars, while delivery, overflow, and opt-outs still need disposable cups, where TakeawayPack's rim-standardized cups, lids, sleeves, and carriers form the other half of a mixed system.

Two programs bracket the outcomes to plan between.

The closed-loop end: Starbucks SODO 8, inside the company's Seattle headquarters. Since early 2022 this café has served every beverage in a reusable cup by default, with returned cups collected at kiosks, professionally cleaned and sanitized, and put back into circulation within about 48 hours. A later Starbucks test at a community-store location inside a Seattle iceplex applied the same model — every hot and cold drink in a returnable cup, bins around the facility, professional washing — in a venue where customers stay nearby while drinking. Starbucks has framed offices, universities, hospitals, and airports as the natural expansion environments precisely because the customer stays inside the loop's geography.

The open-loop end: the Petaluma Reusable Cup Project, California. Over 12 weeks in 2024, 30 businesses across Petaluma swapped single-use cups for reusable purple cups at no charge and no deposit, supported by more than 60 return bins on sidewalks, in parks, on patios, and at drive-thrus. The published results report the numbers that matter to anyone budgeting a system: 51% of cups were returned citywide, 57% in the walkable downtown, over 220,000 cups returned, and a stated environmental break-even point of 40–50% returns depending on the hot/cold cup mix. The report also notes that 79% of returned cups came back somewhere other than where they were purchased, and that the local materials recovery facility recovered roughly 24,000 stray cups from recycling streams.

A campus sits between these poles but closer to the headquarters case: customers mostly stay inside a defined geography on predictable routines — the most favorable condition for reuse economics. The lesson from both programs is that loop geography, return-point density, and the default-versus-opt-in choice move every number below. One honesty note before the math: no public source publishes a cup-level circulation count or loss figure for the Seattle headquarters store, and Starbucks has not disclosed return rates for its market pilots — treat every borrowed number as a hypothesis for your own pilot to confirm.

1. How many cups must be in circulation? Determines fleet size and initial capex. Answer: POS volume × engineered adoption × measured days-out.
2. Where and how many return points? Determines recovery rate, bin capex, and collection labor. Answer: a foot-traffic map of your campus.
3. How fast can cups be washed and redeployed? Determines peak capacity, turnaround, and hygiene compliance. Answer: wash equipment specs or vendor terms.
4. What share of cups never comes back? Determines replacement budget and break-even viability. Answer: tracked inventory counts, not assumptions.

Answer them in that order: fleet size depends on turnaround, turnaround depends on return-point density, and loss rate corrects all three.

The core formula is a flow equation:

Cups in circulation = Daily borrow-cup orders × Average days out of circulation × Safety buffer

Start with drinks per day from your POS, then multiply by the adoption rate you engineer for. Opt-in with no fee and no default yields low adoption — nobody who has to ask will ask often. Default reusable with an easy opt-out shifts most orders to the reusable cup, because the counter handout is the path of least resistance; SODO 8 and Petaluma both ran default models. Default plus a deposit or single-use fee narrows opt-outs further — the 2021 Seattle pilot charged a $1 refundable deposit, repaid as app credit when the cup was scanned back at a kiosk. Model two adoption scenarios and size the fleet for the one you will commit to.

This is the number operators get wrong, because it has four components: customer holding time (handout to bin drop — same-morning returns are the best case; overnight at a desk or a weekend at home is the campus reality); collection delay (how often bins are emptied into the wash stream); wash and sanitize turnaround (in-house can be same-day; the 2021 Seattle pilot's third-party wash ran about 48 hours); and restocking delay (clean cups sitting in a crate instead of at the espresso station). Sum the four, add a safety buffer of 20–50% for peak days, slow returns, and damaged stock pulled mid-cycle.

A conservative campus case: 300 drinks/day from POS; 60% adoption on a default launch gives 180 borrow-cup orders/day; 1.5 days of customer holding (cups sit at desks until the next coffee run) plus 0.5 days for collection, wash, and restock gives 2.0 days out; a 25% safety buffer gives a fleet of 180 × 2.0 × 1.25 ≈ 450 cups.

Run the same math at a 3-day worst-case cycle and you get roughly 675 cups; at a tight 1-day cycle, about 225. That spread — a factor of three from cycle time alone — is why you should buy the conservative number only for the pilot, then resize once you have measured your real days-out. The most valuable pilot output is actual average days out of circulation: it converts your fleet from a guess into an inventory policy.

Two fleet details from the reference programs. Sizes multiply your fleet: Starbucks' returnable cups come in three sizes; Petaluma's multi-brand cups ran 12 oz and 16 oz hot/cold formats plus a 22 oz fountain cup — two borrow sizes means two fleets sized to their own demand, or one borrow size with the other served in single-use. Lids, straws, and sleeves stay consumable even in reuse programs (see Section 7) — budget them inside the program, not as one-time purchases.

The clearest finding across all three programs is that convenience of return drives recovery more than rewards do. At the Starbucks headquarters, return points are distributed throughout the building and kitchen areas, with the stated design goal that returning a cup must be as easy as discarding a single-use one. University campus tests placed bins beside existing trash and recycling stations and where students gather, then expanded based on where cups piled up. In Petaluma, more than half of the 60+ bins sat in the walkable downtown — and the downtown return rate (57%) beat the citywide average (51%). Petaluma ran with no deposits, no fees, no sign-up — automatic opt-in, zero friction — and still cleared break-even. A refundable deposit is loss insurance, not a behavior engine.

Map where drinks finish, not where they are bought. The highest-value spots, in rough order: the café counter (immediate returns); building lobbies and elevator banks (commute nodes everyone passes daily); office kitchenettes and break rooms (where desk cups accumulate — your highest-yield zone in a workplace campus); cafeterias and food courts (lunch traffic carries morning cups); parking exits, shuttle stops, and entrances (catches cups on the way out); and beside existing waste and recycling stations (people already carry empties there).

A practical starting density for a single-café pilot is 2–4 stations (café, main lobby, largest office building, cafeteria), expanded later by observed fill rates. Bin design matters as much as count: Petaluma's bins used a cup-shaped drop-slot silhouette so they could not be mistaken for general trash — reported contamination was minimal — while a generic open bin fills with sandwich wrappers within a week.

Cups returned per station per day (fill rate, rebalancing need); time from purchase to return (your days-out input); and the share of returns at non-purchase locations (validates the placement map).

Decide the washing model before buying cups, because it constrains turnaround, labor, and even cup material choice.

Three models fit a campus. In-house washing suits steady volume with back-of-house space and a commercial dishwasher. A third-party wash operator suits pilots, or where your health authority expects commercial-grade sanitation you cannot host — nail down pickup frequency, minimum batch size, transport cost, and turnaround guarantee. A hybrid suits a rush profile: the café washes urgent morning stock, the vendor takes the overflow.

Peak-hour capacity must exceed peak-hour inflow with margin. If 180 borrow-cup orders return across an 8-hour day but 40% land in the two hours after lunch, your in-house operation must clear roughly 90 cups in two hours at peak — not the 22.5/hour daily average — plus batch capacity to sanitize the morning fleet before the next rush. Batch the work: run 2–3 bulk wash cycles per day (mid-morning, mid-afternoon, close) rather than real-time washing through the morning peak.

Budget collection runs, wash operation, inspection, and restocking as incremental daily labor hours, and carry them into the cost-per-use math in Section 5. Petaluma moved sorting and washing entirely off-site so store staff carried none of it — the report names this as a design goal, and it is a procurement decision: the difference between a loop your baristas run and a loop your vendor runs.

Reuse programs are regulated as food service. Washington State's retail food code (which adopts the FDA Food Code) is the framework the Seattle tests operated under, and equivalents exist in most jurisdictions. Returned containers may be cleaned and refilled only if designed for reuse, initially provided by the establishment for the purpose of being returned, cleaned and sanitized per the code's warewashing requirements, and visually inspected by a food employee before re-entering service (WAC 246-215-03348). Hot-water sanitizing in mechanical warewashing requires a rinse of at least 180°F (82°C) at the manifold for most machines (165°F/74°C for stationary-rack single-temperature machines) and a utensil surface temperature of 160°F (71°C) verified by irreversible registering indicator (WAC 246-215-04555, WAC 246-215-04710). Air-drying is mandatory after cleaning and sanitizing before food contact — cloth drying is prohibited (WAC 246-215-04900).

Three consequences. Cup material must survive commercial wash heat repeatedly — Petaluma selected expanded polypropylene for insulation and strength; whatever you buy, get the rated wash-cycle lifetime in writing and validate it in your pilot, because your amortization depends on it. If you outsource washing, the vendor's process is what your health authority will inspect, so put sanitization method, temperatures, and the inspection step in the contract. And "washed, sanitized, inspected" is your trust story: Petaluma's research found that explaining the cleaning process in those three steps measurably helped people get past cleanliness concerns.

No public source gives a trustworthy, transferable loss number for a campus cup loop, and figures circulating from internal corporate programs are frequently misattributed. Do not budget from anyone else's percentage — including the numbers in this article, which describe other people's systems. Structure your loss accounting and measure it in the pilot.

Customer non-return — cups kept at desks, taken home, or carried out of the campus; Petaluma's focus groups heard people admit to keeping cups for home reuse, and its MRF partner recovered about 24,000 cups from recycling streams. Operational loss — cups lost between bin, transport, wash, and restock, trackable with counts at each handoff. Damage and wear-out — cracks, staining, failed inspection pulls; scheduled replacement, not a loss event.

Serially number or QR-code the fleet and log cups out at handout and in at the wash station — even a clipboard tally by cup series gives days-out and return rate per cohort. Do a weekly physical count: fleet total = clean stock + dirty stock + in-wash + issued; the gap between tracked and physical counts is your loss signal. And watch campus recycling and trash bins the first month — that is where Petaluma's non-returners' cups went.

A closed workplace campus should outperform an open city system: Petaluma's no-deposit citywide loop returned 51% overall and 57% downtown, and a contained campus with default handout and dense bins has structurally better odds, because cups cannot easily leave the geography. That direction of reasoning is sound; any percentage for your site is a guess until measured. Recovery has a break-even threshold, not a perfection requirement — Petaluma put its environmental break-even at 40–50% returns, and your financial break-even is its own number (Section 5). Deposit vs. no-deposit is a real design fork: the 2021 Seattle pilot charged $1 and repaid it as app credit; Petaluma ran deposit-free on saturation and simplicity. Deposits reduce loss exposure but add steps that depress adoption — Petaluma frames this as the inclusivity dilemma: penalties raise return rates among the motivated while excluding everyone else. On a private campus with employee IDs you may have a third option: link borrowing to an account and cap concurrent borrowed cups instead of charging money.

Build the comparison per drink served, not per cup bought:

Cost per reusable use = (cup amortization + washing + collection labor + loss replacement + admin) ÷ total uses served

Cup amortization = purchase price ÷ lifetime washes. A cup that survives 300 washes and costs $3 amortizes to one cent per use before operating costs; a cup that cracks after 40 washes amortizes to 7.5 cents before a single drop of water. This is why the wear data from your pilot matters more than the invoice price of the cups.

Compare against your single-use spend all-in: cup + lid + sleeve + straw + waste hauling per drink. Two honest notes from the reference material: reuse is not automatically cheaper — the Petaluma report states plainly that reuse remained considerably more expensive for businesses than single-use cups at that scale, with cost reduction expected from reverse-logistics scale, lower loss, and wash-process optimization, not from cup price. And the environmental case has a threshold, not a binary verdict: every unreturned cup's production burden is unrecovered, so financial and environmental break-even are different numbers; calculate both.

Spreadsheet lines to build: borrow-cup orders/day; days out of circulation; fleet size (Section 1 formula); cup unit cost ÷ expected washes; wash cost per cup; collection labor hours × wage; monthly replacement = fleet × monthly loss rate; cost per reusable use (sum ÷ uses); current cost per disposable drink (cup + lid + sleeve + straw + hauling); and the break-even return rate solved from the above.

Design: one café, 300–600 cups (your conservative fleet number), 2–4 return stations, serial or QR tracking, default-reusable for stay-on-campus orders with an explicit opt-out. Metrics and thresholds: return rate (wash-station counts ÷ cups issued) must clear your calculated break-even; average days out should land ≤ 2 days; clean stock should never hit zero at open; contamination (trash pulled per 100 returns) must stay low enough that sorting is not a labor sink; wash turnaround must match your fleet-math cycle assumption; and inspection failures (cups pulled per 100 washed) reveal the real cup lifetime for amortization.

Questions for a wash vendor or equipment supplier: What sanitization method and verified temperatures do you run? What turnaround do you guarantee, and what happens when a peak day exceeds capacity? What are your minimum batch size, pickup frequency, and transport cost per cup? What is your inspection standard, and who pulls a damaged cup? Which cup materials have you validated, at what observed wash count before retirement? Can you report per-cup scan data so I can compute days-out and loss rates without building my own system?

A campus loop has a natural addressable base — the stay-on-campus regulars who buy at the counter and finish their drink inside the geography. It does not cover everything, and pretending otherwise is how pilots fail on their first delivery rush. Delivery and takeaway beyond the campus need single-use packaging that travels. Peak overflow happens when the borrow fleet is in circulation or in the wash; the answer is disposables, not turning customers away. Opt-outs and one-off visitors always need a disposable option — even motivated participants cannot return every cup every time. And lids, straws, and sleeves stay consumable even in Starbucks' own tests — only the cup body loops.

So the procurement reality is two systems specified to work together: a reuse loop sized by the arithmetic above, and a single-use stock covering the remaining volume without colliding with it at the counter.

That hybrid is the part TakeawayPack supports directly. TakeawayPack is a foodservice packaging supply chain partner for international buyers — with its own printing operation working alongside a network of manufacturing partners — supplying the disposable side: cup and lid families built around standardized rims so one lid format covers multiple cup sizes, cold-drink PET cups in the 90/93/98 mm rim families with matching dome lids, cup sleeves and four-cup carriers for hot-drink handoffs, and single-cup and double-cup bags for carry-out. The same rim-first logic keeps the overflow path painless: fewer lid SKUs alongside the reuse program, fewer wrong-format surprises during a rush. To be clear about the boundary: TakeawayPack does not supply reusable cup systems, wash equipment, or return infrastructure — the loop itself needs a cup vendor and wash partner built for that job.

If you are specifying that disposable half, TakeawayPack works from product category, material, size or capacity, expected quantity, customization needs, and destination market — and custom printing runs on a digital proof and dieline simulation confirmed with you in writing before production starts, the same "measure before you commit" discipline this article applies to cup fleets. MOQs depend on product and production method: as a general frame, customized standard items typically start around 1,000 units and bag products around 5,000, with the final number confirmed in the quotation.

1. Pull 30 days of POS data: drinks/day, share by size, share that stay on campus vs. leave.
2. Build the fleet model from Section 1 with two adoption scenarios; write down your assumed days-out so the pilot can prove it right or wrong.
3. Walk the campus and mark every point a person passes with an empty cup: lobby, elevator, kitchenette, cafeteria, parking exit. Shortlist 2–4 return-station locations.
4. Check your local retail food code for returned-container cleaning, sanitization, and inspection requirements — Washington's WAC 246-215 is a worked example of the structure you will find.
5. Get wash quotes (in-house equipment and one third-party operator) using the vendor questions above, and price cups rated for commercial wash cycles.
6. Decide the incentive fork — deposit, no-deposit default, or account-linked cap — and note that Petaluma cleared break-even with zero friction, so simplicity is a legitimate strategy.
7. Keep the disposable half under active cost review; it will carry your delivery, overflow, and opt-out volume for the foreseeable future. Browse disposable cups and lids for rim-standardized options, and see the total-cost buying framework for sustainable takeaway packaging, the reusable-cup workflow design guide for mobile orders and drive-thru, the rim-first cup program compatibility map, the cup, lid, and drink matching guide, and how to measure whether a packaging upgrade pays back. When you are ready to specify the disposable side, contact TakeawayPack with your volumes and destination market.

• ESG Today, Starbucks Tests 100% Reusable Cup Store (2022)
• NextGen Consortium / Closed Loop Partners, Petaluma Reusable Cup Project results report and results announcement
• Washington State Legislature: WAC 246-215-03348, WAC 246-215-04555, WAC 246-215-04710, WAC 246-215-04900