The Worst Day of the Year

How a theme park is designed to handle its hardest moments

There is one question almost nobody asks when talking about theme parks: how many people can this park absorb before the experience falls apart?

Not how many rides it has. Not how many square metres. Not what records it breaks. How many people, at what time of day, and what happens when that number is exceeded.

This chapter is an engineering capacity exercise applied to the project. The premise is straightforward: if you don’t know how many visitors you can serve before the experience degrades, you are designing blind.

Photo by Lisa from Pexels from Pexels: https://www.pexels.com/photo/group-of-people-walking-in-a-parade-16845591/

The design day

A park targeting one million visitors per year, operating 280 days annually, receives an average of 3,571 visitors per day. That is a manageable number.

The problem is that demand does not distribute evenly. A realistic estimate of the annual spread:

19.5% of the year’s visitors arrive on 5% of the days. This is not a quirk of this project — it is the standard pattern for any European theme park with seasonal weather.

The number that sizes all the infrastructure is not the average: it is the peak. A park is designed for its worst day, because that is when the promise to the visitor is actually tested. If the system handles 15,000 people with a reasonably good experience, the rest of the year is straightforward.

For this project, the design peak is 15,000 visitors in a single day.

The first misconception: nobody is inside at the same time

There is an important difference between daily visitors and people simultaneously in the park. The arrival curve follows an approximately normal distribution: most visitors enter between 10:00 and 12:00, the bulk are inside between 12:00 and 17:00, and departures spread gradually from 18:00 until closing.

At peak pressure — typically between 13:00 and 16:00 — the simultaneous occupancy factor sits between 60% and 70% of the daily total.

With 15,000 visitors:

Maximum simultaneous occupancy = 15,000 × 0.65 = 9,750 ≈ 10,000 people

The park must be dimensioned for 10,000 people at the same moment, not 15,000.

How much space does that imply? Assuming 35 hectares of park area and that 47% of that surface is usable by visitors (the rest is service zones, backstage, dense vegetation, machinery):

Usable area = 35 ha × 0.47 = 164,500 m²
Peak density = 164,500 / 10,000 = 16.45 m²/person

To calibrate that number: below 5 m² per person the experience feels saturated and uncomfortable. Between 10 and 20 m² is the typical range for European reference parks. Efteling operates habitually between 12 and 18 m² per person. 16.45 m² at absolute peak is comfortable density.

The 35-hectare park does not enter spatial saturation under normal design conditions. The problem is not space — it is the processing capacity of the attractions.

The formula that decides everything

Every attraction has a critical parameter: the number of people it can serve per hour. In the industry this is called CPH (Capacity Per Hour) or throughput.

The general formula is:

CPH = (60 / cycle_time) × seats × vehicles × η

Where cycle_time is the total cycle time in minutes (ride plus loading and unloading), seats and vehicles define the physical configuration, and η is the real utilisation factor (between 0.80 and 0.92 depending on attraction type — no system operates at 100%).

Applied to specific formats:

The bottom of that last range is not a typo. An attraction with 8–10 minute sessions and fewer than 25 riders at a time serves under 200 visitors per hour. The flying theater serves six to eight times more in the same period. No amount of theming or design changes that arithmetic.

The mathematics of the queue

If 15,000 visitors each want to ride every attraction once across the day, and 60% of that demand concentrates in the six peak hours (12:00–18:00), the hourly demand arriving at each attraction during that window is:

Average peak demand = (15,000 × 0.70 × 1.6) / 12 ≈ 1,400 PPH per attraction

The 0.70 factor assumes not every visitor attempts every attraction, and 1.6 is the hourly concentration factor relative to the daily average.

A simplified queuing model (D/D/1) shows what happens when demand exceeds capacity:

The bottom two cases deserve careful reading. A 160-minute queue is bad but manageable; a queue of more than 16 hours is not a queue — it is an attraction that, in practice, completes only three or four useful cycles during the entire day. It cannot operate openly with 15,000 visitors in the park without active demand management: mandatory Virtual Queue, time-slot booking, or both.

It is worth separating two problems that the model tends to conflate. A low-CPH attraction with a long queue simply draws a portion of visitors who, without that attraction, would be distributing themselves across the rest of the park. In that sense, even low-throughput attractions have a distributing effect: they do not absorb much demand, but whatever they absorb stops pressing on other queues. The real issue is not that they generate long queues for themselves — it is that, from the park’s aggregate capacity perspective, they contribute almost nothing. Adding a 150-CPH attraction when the portfolio already has a throughput deficit is like fitting a tap where you needed a hose. The visitor appreciates having it; the park’s absorption capacity does not improve.

How many attractions, and what kind

There is a straightforward balance equation: the sum of all attraction CPH values must equal or exceed the aggregate peak-hour demand.

For this park, fixed-capacity shows (a 450-seat hall running 5 sessions per day and a 280-seat theatre running 6) absorb roughly 380 additional visitors per hour at peak. The mechanical attractions must cover the rest:

Demand mechanical attractions must handle: ~11,620 PPH aggregate

If the portfolio averages a CPH of 900 (a reasonable minimum):

Minimum mechanical attractions = 11,620 / 900 = 12.9 → 13

With the two shows: minimum 15 total experiences in Phase 1. That is exactly the number in this project’s proposed portfolio, with no margin to spare.

One implication follows directly from this equation: adding low-CPH attractions does not improve the park’s capacity — it dilutes it. A park with 20 attractions, eight of which have CPH below 400, can have worse aggregate capacity than a park with 15 well-dimensioned attractions. The number of attractions is not the right indicator. Aggregate CPH is.

The problems nobody sees but the park feels

Attractions are the visible problem. The capacity model has other implications that are equally critical and rarely appear in conversations about theme parks.

Parking. With 15,000 visitors, assuming 15% arrive by public transport (the target with a shuttle from the regional rail network) and average vehicle occupancy of 2.8 people:

Vehicles on peak day = (15,000 × 0.85) / 2.8 ≈ 4,554 cars

A 4,000-space car park — the initial figure in the model — runs a deficit of 554 spaces on the absolute peak day. The solution is not an unlimited car park: it is reserving land for 4,500 spaces and raising the public transport target to 20%. With that adjustment, peak parking demand drops to 4,286 vehicles, and the margin holds.

Food service. During the lunch peak (13:00–15:30), roughly 35% of visitors look for a restaurant seat at the same time. With 15,000 visitors and an average meal duration of 40 minutes:

Simultaneous diners at peak = 15,000 × 0.35 × (40/180) ≈ 1,167

The park needs between 1,200 and 1,400 distributed dining seats to prevent the lunch hour from becoming another queuing problem. Not one central canteen — multiple points spread across the park to decompress flow at critical hours.

Staff. The standard ratio at quality European parks is one employee for every 15–20 simultaneous visitors. With 10,000 people in the park at the same time:

Peak shift staff = 10,000 / 17 ≈ 590 people

590 people on a single work shift. That is a medium-sized company in its own right, with everything that implies in terms of management, training, turnover, payroll and internal logistics. Running a theme park on a peak day is not an event — it is an industry operating in real time.

What the model says

The capacity model produces several results directly usable in the design.

The 15,000-visitor peak generates 10,000 simultaneous people in the park. The resulting density (16.45 m²/person) is comfortable spatially — the park does not saturate physically, it saturates in queues.

A CPH of 900 is the threshold below which an attraction requires active demand management. Low-throughput attractions — junior rides, long-cycle formats — are not design errors: they are decisions with specific operational consequences that must be made with open eyes.

The 15-experience portfolio produces an aggregate CPH of approximately 10,460 PPH from mechanical attractions. The deficit against the target (11,620 PPH) is around 10%, recoverable through Virtual Queue on low-throughput attractions and deliberate show scheduling during peak pressure hours.

Parking, food service and staffing are real bottlenecks, not secondary considerations. A park that manages queues well but fails on parking has exactly the same experiential problem as one that manages queues badly.

The next chapter develops the financial model: what investment this requires, what revenues it can generate, and where the project becomes viable — or does not.

This series examines the feasibility of a theme park in Asturias from scratch, using the same tools that would apply to a real project. The models are estimates; the conclusions are for the reader to draw.

Published chapters:

• Chapter 1 — Could Asturias Have the Best Theme Park in Spain?
• Chapter 2 — Where to Build It
• Chapter 3 — What Kind of Park Should This Be (publishing soon)
• Chapter 4 — The Worst Day of the Year (this article)