When workers arrive all at once at a set time, such as before a shift starts, before vehicles are dispatched, or before a work crew assembles, the alcohol check queue can grow long. Safety and health or general affairs staff must run the checks so that everyone finishes before work starts, but the daily headcount alone makes it hard to decide how many test stations and guides are needed. Even with the same 100 people, the queue differs between arrivals spread evenly over an hour and most people arriving in the first 15 minutes.
An alcohol check here means the operating procedure in which workers arriving for work, or people about to start a task, have their alcohol status checked by a breath test or similar method and leave the check area after the result is confirmed. Throughput should be calculated not as the number of tests one breathalyzer can perform on paper but as the number of people who have completed everything from identity check through guidance, testing, result confirmation, and moving on. If you compare arrivals and actual completions during the peak in the same time units, you can judge when the queue builds up and when it clears.
Break the arrival peak into short intervals
First, divide the time from when checks start until the queue disappears into short, equal intervals. Ten-minute intervals are not always right, so choose a unit that fits the site and can show differences in how arrivals bunch up. Sites where shifts arrive exactly on the hour can use shorter intervals, while sites with more spread-out arrival times can use somewhat longer ones.
For each interval, record three things: the number of people who newly joined the check line, the number who finished the check and left the line within that interval, and the number still waiting at the end of the interval. If contractor staff or other crews use the same entrance, the arrival count must include everyone who actually joins the check line.
| Figure to record per interval | What you actually count | Why it is needed |
|---|---|---|
| Arrivals | People who newly joined the check queue | Shows how concentrated arrivals are |
| Completions | People who finished through result confirmation and left the queue | Reflects the processing capacity of the whole procedure on site |
| Remaining queue | People still waiting at the end of the interval | Shows whether delays carry over to the next interval |
| Separately handled | People taken out of the normal line for retests, guidance, or error handling | Separates out the causes that slowed the normal queue |
Carry the remaining count into the next interval
The remaining queue is found by adding the people left from the previous interval to this interval's arrivals and subtracting the people who completed in this interval. If the result is less than 0, treat the remaining count as 0. This method keeps you from missing congestion right before work starts, even when the total headcount and the day's completions are equal.
The key judgment criterion is not whether a queue forms in the busiest interval but whether that queue disappears before the set start time. Even if arrivals drop in the next interval, the line stays if completions are not enough. Conversely, if a short queue forms but clears in a later, quieter interval, you can compare options such as adjusting the check start time or the guidance route before adding test stations.
Read queue buildup with a hypothetical case
Consider a hypothetical case to illustrate the calculation. Assume a check point can complete checks for 20 people every 10 minutes, and that 30 people arrive in the first 10 minutes, 20 in the next 10 minutes, and 10 in the third 10 minutes. In the first interval, 20 people complete and 10 remain. In the second interval, of the 30 people in total, the 10 left over plus the 20 new arrivals, only 20 complete, so 10 remain again.
In the third interval, all 20 people, the 10 remaining plus the 10 new arrivals, complete and the queue disappears. The observed result is that a remaining queue of 10 people persists for the first 20 minutes. This does not mean the total daily processing capacity is insufficient; it leads to the interpretation that completion capacity in the first two intervals, when arrivals are concentrated, is lower than the arrival rate.
Possible actions you can pilot include moving up the check start time, splitting identity checks and test guidance between different people, or changing the route so people wait in a single line and then move to whichever test station is free. If you consider an additional test station, also look at whether the person guiding and confirming results at that station, the waiting space, and the flow of consumable preparation change too. The completion criterion in this case is that the normal queue has 0 people remaining at the set start time and that, even when retests occur, delays in the normal line do not carry over to the next interval.
Separate test time from operational completion
The time for one test to finish on the breathalyzer is only part of throughput. If identity checks are slow, consumables are not ready in time, or people who have seen their result linger in the passage because they cannot find where to go next, the test station sits idle. So it is more suitable for operations planning to look at valid completions through result confirmation rather than the number of test attempts.
Count failed attempts and retests separately rather than mixing them into completions. For example, if completions were low in a particular interval but the test station was often empty, that is grounds to look first at the guidance, verification, and movement flow rather than the test itself. Conversely, if the queue in front of the station was long and time was concentrated on retests or result confirmation, you can consider setting up a separate guidance location and handling flow apart from the normal queue.
If a line remains when observation ends, keep recording until everyone in it has finished the check. If you leave out people processed after work starts, the planned margin can look larger than it really is. Especially for work where the start time matters, such as work assignments or vehicle operation, you need to record the time the queue clears as a result just as important as the number of completions.
Compare before and after adding stations under the same conditions
Even if you add test stations, completions may not rise in the same proportion if one operator must handle identity checks, consumable preparation, and result confirmation across several stations. A single line feeding whichever station is free and separate lines per station can also produce different results depending on the actual arrival pattern. Do not compare only the number of devices; compare how completions and remaining queues per interval change under each approach.
The figures used in the calculation should come from observations on site rather than the conditions in product brochures. Record arrivals, completions, remaining queues, and the flow of retests and guidance on the day with the heaviest arrivals or under similar conditions, then apply the change and look again in the same way. Article 36 of Korea's Occupational Safety and Health Act sets out risk assessment: identifying hazards and risk factors at the worksite and establishing and implementing improvement measures to reduce the risks. A flow in which arrival congestion causes checks to overlap with the start of work can also be examined in connection with site risk factors, and after improvement you can confirm as an operating result whether the queue clears by the set time.
