Key takeaways
- Divide each compatible demand pool by what one trip can actually carry and round up. The result is a capacity lower bound; stop sizes, reset locations, outlet rules, and the working clock may require more trips.
- Usable trip capacity is whichever ceiling gives out first: tank or body volume, legal payload, or your buffer-adjusted operating limit. Federal weight caps can end a load while the tank still looks half empty.
- Every trip ends at a reset: a disposal gate, transfer station, depot, refill point, or washout. Each reset carries its own hours, queue, fees, and accepted materials.
- Stops join trips as geographic clusters that finish near their reset, and time windows, material compatibility, and outlet acceptance pull the exceptions.
- A bigger truck beats a multi-trip day only when the reset loops it removes cost more than the truck adds in ownership, access, and weight risk. When the clock breaks first, compare a second truck, allowed demand shifts, and other feasible capacity options.
- The planned trip count is a forecast. Scale tickets, meter readings, or tank measurements recorded at resets can reprice the trips still ahead.
Multi-trip route planning treats the second and third loads as part of the day's design rather than as whatever happens after the truck fills. We've covered what capacity-based planning is and the mid-route math for a truck that fills early; this article is for the day you know in advance is bigger than the truck. I build DynoRoute, routing and dispatch software for fleets whose trucks fill up or empty out as they work, and I've spent this year interviewing the operators who run them. The pattern that repeats: trip one is planned, and every trip after it is improvised. What follows is the arithmetic and the placement rules that make trips two and three as deliberate as trip one.
How many trips does the day physically need?
The capacity lower bound is one line of arithmetic: capacity-bound trips = total compatible demand ÷ usable trip capacity, rounded up to the next whole trip. It is the fewest trips the load could occupy if it can be partitioned efficiently. Indivisible stops, separate material pools, outlet acceptance, reset placement, time windows, and shift limits can push the feasible plan above that number, so sequencing still has to prove it.
Both inputs need honest numbers. Total compatible demand is the sum of every stop's expected load that can legally and physically share a tank or body. Streams that cannot mix are separate pools: septage and hydrocarbon waste may share a calendar, never a compartment, so each pool runs its own lower-bound math. Estimate each stop from a relevant distribution of recent comparable actuals where possible, adjusted for known changes; then use measured contracted quantity or a conservative default when history is not available. One last actual is useful evidence, not automatically the best forecast.
Usable trip capacity is whichever limit gives out first, minus your buffer. Volume is the visible ceiling. Weight hides: federal interstate rules cap a vehicle at 80,000 pounds gross, 20,000 on a single axle, and 34,000 on a tandem, and dense material can put an axle over while the sight glass reads half full. Volume capacity is not legal payload. Compute the ceiling in whichever unit binds for your material, and hold back a buffer sized from your own overruns rather than a borrowed percentage.
Here is the shape of it, with invented round numbers. Tomorrow's compatible demand across a territory sums to 8,200 gallons. The truck carries a 4,000-gallon tank, the illustrative operating plan holds it to 90%, or 3,600 gallons, and against this material's density the scale gives out before the tank does, at 3,000 usable gallons a trip. 8,200 ÷ 3,000 is 2.73, so capacity alone requires at least three trips. Even if 3,600 gallons were usable, 8,200 ÷ 3,600 is 2.28: still a lower bound of three. Small capacity changes often leave the rounded result unchanged; compute the bound first, then prove the feasible trip plan.
Read the decimal before rounding it away, because the fraction is the plan's slack. A 2.73 means the last trip runs at about three-quarters load, with room for estimate error. A 2.05 also rounds to three, but now the day pays a full reset loop to move one stop's worth of material; those stops are candidates for another truck or another day. A 2.95 is three trips with no slack at all, one heavy stop from becoming four.
Operations research calls the basic capacity-constrained version the capacitated vehicle routing problem: demand assigned to vehicles so that none exceeds its capacity. A multi-trip day is a coupled extension in which the same truck can serve several loads and each reset consumes time and changes what remains feasible. It is not several independent single-load solutions, because the trips share one vehicle, one shift, and linked reset choices. A tool that models only one load per truck per day has no place to represent that coupling.
Where does each trip reset?
Every trip ends at a reset: the fixed facility where the truck becomes useful again. What kind of reset it is, and where it sits, shapes the trip before it and the trip after it, because trip one's reset is trip two's starting line.
"Reset" is deliberately broader than "dump," because fleets that fill up and fleets that empty out both run multi-trip days:
| Reset point | What the truck gets back | What the plan must carry |
|---|---|---|
| Disposal or receiving facility | An empty tank or body | Gate hours and last-load cutoff, queue, tip fee, accepted materials |
| Transfer station | An empty body, closer to the route | Its own fees and acceptance rules against the drive it saves |
| Depot or yard | Fresh containers, empties dropped | Loading time and container inventory on hand |
| Water or product refill | A full tank to work from | Fill time, source hours, availability |
| Washout | A clean tank between incompatible loads | Time and water that collect nothing but make the next trip possible |
Position comes first. A reset that borders the next trip's cluster turns the return leg into progress; a reset across town from everything charges the day a deadhead each way, every trip. The geometry compounds at fleet scale: a 2023 study modeling municipal waste collection in Bahía Blanca, Argentina concluded that installing a transfer station was worthwhile on traveled-distance grounds. You cannot move the disposal plant, but you choose which facility each trip ends at and which cluster sits in front of it. Every reset is a route decision, not a fact of geography.
Each reset also has a running cost the plan must budget: travel out, queue, unload, cleanup, travel back, plus the fee. That anatomy, and the fee spreads that make facility shopping worth it, is the subject of route planning with disposal stops.
Two resets require explicit policy. Decide whether the last trip must end with a final offload based on tomorrow's capacity needs, the receiving site's hours, and the rules for storing that material overnight. Between incompatible loads, schedule whatever cleaning or washout the equipment, outlet, and applicable rules require. The mistake is not choosing one policy or the other; it is leaving the reset out of the route while assuming the driver will absorb it.
Which stops ride which trip?
Assign stops to trips as geographic clusters that each finish near their reset, then let three constraints pull the exceptions: time windows, material compatibility, and what each outlet will accept.
The common failure is slicing instead of clustering. Sequence the whole day as one drive-time-optimal chain, run it until the truck fills, call whatever remains trip two. The slice point lands wherever the arithmetic happens to put it, mid-cluster and far from any gate. The truck deadheads to the reset, deadheads back, and the afternoon crosses its own morning tracks. Trips built as clusters, each with a reset on its border, spend those same minutes collecting. The corridor logic is the one that governs route density: stops close together on the same trip are nearly free, and the expensive miles are the ones between clusters.
Then the exceptions, in order of how hard they pull. Time windows pin stops outright: the account that must be pumped before it opens at 9 a.m. rides trip one whatever cluster it sits in, and a facility with a 2 p.m. last load drags its whole trip earlier in the day. Material compatibility decides what may share a trip, since incompatible streams can share a day but never a tank. The outlet closes the loop: a trip must end at a gate that accepts what that trip collected, at an hour it is willing to take it. A perfect cluster attached to the wrong facility is still a broken trip.
One quieter rule earns its keep: put the highest-uncertainty stops on the earliest trip you can. New accounts, seasonal accounts, stops whose last three actuals disagree — front-load them. The reason is the repricing covered below: the plan improves every time a trip closes, but only for the stops still ahead of it. Front-loaded, that uncertainty turns into information while there is still a day left to use it.
One bigger truck, a second truck, or a multi-trip day?
A multi-trip day is often the cheapest scenario to test because it uses the truck already assigned, but it is not automatically the cheapest plan to run. A bigger truck wins when the reset loops it removes cost more than the truck adds. A second truck wins when the clock, rather than capacity, is what breaks.
Run the bigger-truck case as a yearly comparison. The saving is knowable: loops eliminated per day, times the loop's time-and-miles cost, times route-days per year. Quantity-based disposal fees may stay similar, while per-load charges and minimums may fall with fewer trips; use the facility's actual fee basis. Against that saving sit purchase or lease, insurance, and two costs that hide until the truck arrives. Access is one: the bigger body has to fit the same alleys, culs-de-sac, and weight-posted roads the smaller one served. Weight is the sharper trap: federal gross and axle caps do not grow with your tank, so added volume can simply move the binding limit to the scale. A truck that carries twice the gallons but, say, only 30% more legal pounds delivers the 30%, and the bigger the body, the more often weight is the ceiling that ends the trip.
The clock can cap trips before the capacity math does. Under federal hours-of-service rules, a property-carrying driver gets up to 11 driving hours inside a 14-hour on-duty window after ten consecutive hours off, where federal rules apply; many local operations run intrastate under state versions, so know which clock governs your trucks. Every reset loop consumes part of that window without collecting another stop. Derive the practical trip cap from your own route and facility history: subtract collection time, breaks, and closeout work from the governing shift window, then divide the remainder by the observed reset-loop time and leave a variance reserve. On grease routes, how many grease traps fit in a working day runs the day-count version of this arithmetic.
So the sequence is: compute the trip count, test it against the driver's window and the facilities' hours, and read the answer per route-day. Three trips that fit the clock is a multi-trip day. Four trips against a three-trip clock is a second truck on that route-day, or demand shifted across the week where the service rules allow it. A trip count that repeatedly breaks the clock, on routes a bigger body can physically reach, is the case where the bigger-truck math deserves a serious run. The same fleet can land on different answers on different days, and should.
Let each reset reprice the rest of the day
A trip count is a forecast, and many reset workflows produce a measurement—a scale ticket, meter read, or tank gauge—while the truck is stationary and the plan can still change. A three-trip day may provide two natural checkpoints if the operation records them; the measurements still require a defined capture process.
The arithmetic is short enough to run from the scale house; the numbers here are invented to show the shape. The plan said 8,200 gallons at 3,000 usable per trip — 2.73, three trips. Trip one was planned at 2,900 gallons and the ticket reads 3,150, so stops are running about 9% heavy. Apply that to the 5,300 planned gallons remaining and you now expect roughly 5,780, which is 1.93 trips against the two you have left. Still feasible, but the slack that was a quarter-load at 6 a.m. is a couple of stops wide at 11 a.m. The decision that follows — trim two stops to tomorrow, or hand the far cluster to the truck running light — gets made at the reset, with a whole afternoon left to absorb it. The same drift discovered at 4:30 p.m. is a missed stop and an apology call.
This mid-day repricing is why the reset should exist as scheduled work rather than remain an instruction in the driver's head. The capacity and workload figures can then be updated, and the remaining jobs reassessed against the real day. In DynoRoute, capacity inputs, live visit status, AI assignment suggestions, conflict checks, and recovery agents provide documented pieces of that loop; representing the reset and its actual quantity is a configured workflow to prove before relying on it.
Choosing software for multi-trip days
Whatever tool you use, hold it to five requirements before trusting it with a multi-trip day. It should carry each truck's capacity in the units the operation uses and make clear whether the constraint is enforced or merely recorded. It should represent resets as scheduled work with durations, hours, and acceptance rules. It should show load or trip boundaries before dispatch. It should support reassignment or resequencing when reset-time actuals change the forecast. And it should preserve quantities, tickets, photos, notes, and outcomes on the relevant records so next week's estimates can use this week's evidence.
DynoRoute documents per-vehicle capacity inputs, recurring jobs and visits, AI job-to-truck recommendations using capacity, location, availability, workload, travel time, and other constraints, plus conflict checks and recovery agents for disrupted work. Drivers can attach timestamped photos, signatures, notes, forms, and outcome details, including while offline. Those capabilities can support a multi-trip process, but native load boundaries, automatic disposal or refill placement, reset actuals, and multi-dimensional capacity behavior must be demonstrated with your configuration. Plans are published at $199, $499, and $999 a month with credit-based optimization and no per-seat pricing; included credits and plan limits determine which tier fits.
If your dispatchers are counting trips in their heads and improvising trip two, book an intro call and plan next week as trips on purpose.
