National holidays cost the United States 325 billion every year

In the UK we’ve just enjoyed a long four day weekend thanks to two bank holidays to celebrate the Queens Platinum Jubilee.

My colleagues in the Netherlands will be enjoying a day off on the day this post goes out. Another national holiday.

It got me thinking; what countries have the most national holidays? And what is the economic impact of this time off?


National Holidays by Country

National holidays per year by country (2022)

Download chart.

Country Number of public holidays
Mexico 8
Taiwan 9
Brazil 9
Iran 26
Nepal 30
Myanmar 32

Full rankings.

The average number of national holidays across the 67 countries considered is 15. Myanmar has 32 — that’s more than one month of holidays!

National Holiday and affect on productivity

Lost GDP per day (billions) at 50% downtime by Country

Download chart.

Country Number of public holidays Lost GDP per day (billions) at 100% downtime Lost GDP per day (billions) at 50% downtime
United States 11 59.15 29.58
Japan 16 14.51 7.26
Germany 13 10.81 5.41
Kosovo 11 0.022 0.011
Liechtenstein 22 0.020 0.010
Fiji 10 0.012 0.006

Full rankings.

Here the figures assume two scenarios

  1. that during a lost day to a national holiday, output is stopped completely (downtime = 100%) — which is obviously not the case
  2. output is reduced by 50% — probably more accurate

If we assume output drops by 50% on a national holiday, the US still forfeits over $29.5 billion per day! Assuming all national holidays equate to working days off, thats over $325 billion of potential output being lost!

National Holiday and working hours

Actual employment rates are hard to measure (unemployment rate in many countries only captures those of working age seeking a job — not those intentionally unemployed). That being said, it gives us a figure to go off to estimate lost working hours.

Similarly, not all jobs will be 8 hours (what I’ve assumed), nor will all workers get national holidays off.

Lost working hours (assumes ave 8 hrs) by country (2022)

Download chart.

Country Number of public holidays Lost working hours (assumes ave 8 hrs)
India 21 230217120000
Indonesia 16 30561984000
Bangladesh 22 28552392000
Estonia 10 102327280
Fiji 10 67782400
Liechtenstein 22 6407104

In the worst case, that’s two hundred thirty billion two hundred seventeen million one hundred twenty thousand (230,217,120,000) working hours Indians collectively enjoy on national holidays. This is definitely an over-estimate of actual hours, but assuming 50% of this assumption gives a total 115 billion national holiday hours for workers. This equates to about $80 billion in GDP (assuming a 50% reduction in output on these days).

And finally…

As you already know, this post is nowhere near accurate. I am a proponent for more downtime not less. In many studies, a shorter working week produces the same level output, therefore enjoy your holidays everyone!


I compiled an incomplete list of national holidays by country so an obvious improvement would be to obtain a more comprehensive (and validated list). I’ve also made some huge assumptions, (1 national holiday is equal to 1 day off work, 8 hour working day, etc).


I estimate that Indians collectively enjoy over 115 billion national holiday hours (21 national holidays) resulting in an estimate $80 billion USD hit to the countries GDP.

Get the data

  1. Data sources + data used in this post.

Marriott Spends Up To $90 Million a Year on Bathroom Soaps

Save our planet. Please reuse your towel.

For as long as I’ve been staying in hotels as an adult, over 12 years now, almost all hotels ask you to reuse your towel (and they should).

This is one of those small things that when you start to really think about it, as I recently have, leads you down a rabbit hole.

Some hotels have over 7000 rooms (they do exist) — that’s a lot of towels.

And then there are the little bars of soap in the bathroom. You can’t reuse those easily. For a nights stay they might only get used a handful of times.

Examining this at a macro-scale, the numbers are quite astounding (and not in a good way).


Reassuringly, I’m not alone in wanting to learn more about this area of hospitality. In Chekitan Dev and Prateek Kumar from the School of Hospitality Administration (Boston University) conducted a study titled; A Detailed Study of the Expected and Actual Use of Hotel Amenities, that I have used as the foundation for this post.

For all other calculations I used publicly accessible data to calculate the estimates.


Amenity use by hotel guests (2019)

Amenity use by hotel guests (2019)

Download graph.

Actual %
Auto check in 1.00%
In room fitness equipment 1.67%
Restaurant dinner 2.33%
Packaged soap 86.67%
TV 87.33%
Closet 88.00%

View full table.

Auto checkin isn’t really taking off looking at these numbers, nor is restaurant dining at hotels (the latter surprises me).

The majority of guests, 86.67%, all use the packaged soap placed in the rooms bathroom.

Towels are not listed as amenities, but I’d assume their usage amongst guests to be almost 100%.

Number of Hotel Rooms (top 15 hotels by rooms) (2019)

The number of worldwide hotel rooms varies, due to differences in classification. I was however able to find some good numbers quoted for the 15 leading hotels (by total number of rooms).

Number of Hotel Rooms (2019)

Download graph.

Worldwide, Marriott has by far the largest number of hotel rooms, 1.43 million, almost 400,000 more than the second largest hotel chain by room number, Hilton, with 1.04 million.

In total, the top 15 hotels operate a total of 6.33 million hotel rooms.

According to most sources, average occupancy rates at hotels lie between 65% and 80% worldwide.

At 65% occupancy, total occupied hotel rooms amongst these brands is equal to about 4.11 million rooms occupied per night. At 80%, occupancy is 5.06 million. And at 72.5%, occupancy is roughly 4.59 million.

Laundry, at scale

According to TUI:

Every 10kg towel wash consumes at least 50 litres of water and 1.2 kilowatt-hours of electricity, so 3 per cent saves 129,000 litres of water, which means CO₂ emissions for the hotel can be cut by 1,676 kg.

Bath towels are about 0.75kg each. So let’s assume 1kg when slightly wet.

Using the Marriott as an example, at a 72.5% occupancy rate this would mean about 1.04 million kilograms of towels might need to be washed per day. Using the figures above this would require 5.18 million litres of water (enough to fill 2 2.5 million litre Olympic swimming pools) and 124k kilowatt-hours of electricity!

For all 15 hotels that’s 22.93 million litres of water per day (8.37 billion litres per year) and 550,291 kwh of electricity per day (200.9 million kwh per year).

In the UK (according to the UK Government reporting, and thus specific to the UK energy grid):

0.233 kg of CO2e per kWh of electricity

Using this estimation, a rough estimate is that Marriott emits 28,970 kg (28.97 tons) of CO2 everyday just to wash towels.

CO2 Emissions amongst the top 15 chains listed calculates co2 emissions to wash towels to be 128,217 kg per day (or 46.8 million tons of co2 per year)!

I suspect 30%-40% of towels are reused per day (so 70%-60% need to be washed). Assuming the best case, that 40% of towels are reused, it would require 5.02 billion litres of water per year and cause 28.08 million tons of co2 emissions per year.

Throwaway soap

Using the same occupancy rate as before as well as the assumption that 86.67% of guest use soap; Marriott replaces 898k bars of used soap a day, and almost 4 million bars of soap are replaced each day across the 15 hotel chains.

Yearly that’s 1.45 billion bars of soap replaced by these hotels!

These figures are probably an underestimation because I only counted per room. If more than one person is staying in a room, it is likely the second person will open their own bar of soap. However, I will ignore this in my analysis.

Hotels take great pride in selecting their toiletries, and invest a tremendous amount of time and money into finding the right brand partners, and they are  not paying retail prices at this scale. Retail is obviously much higher (e.g. $6 for just 10 soaps here at the time of writing).

Let’s assume each bar of soap costs the hotel between $0.05 and $0.10 at this scale (I couldn’t find any particularly citable cost references here). That’s about $200k and $400k respectively spent by these 15 brands each day on replacement soap (or $72 million and $145 million yearly).

The environmental cost of this one use amenity is also staggering. Assuming the average bar weighs 15g in packaging, and let’s assume 14g when opened and used once or twice, gives a figure of about 56,642 kg of soap thrown away per day (20.31 million kg’s or 20,310 tons of soap a year)  — that’s a lot of landfill.


Whilst almost all the numbers used in this post are estimates, it would be interesting to consider these estimates on a global scale. If the top 15 hotels run a total of 6.33 million hotel rooms, there must be significantly more when all hotels are considered (I’d conservatively estimate at least 4 times the estimates in this post).


I estimate over $145 million could be spent yearly on soap by the 15 largest hotel chains (by number of rooms). That buys them 1.45 billion bars of soap — that are probably only ever used once — weighing 20.31 million kilograms.


  1. Data sources + data used in this post.

Ghost flights from London Heathrow cost airlines almost $200 million in fuel during the pandemic

A landing slot, takeoff slot, or airport slot is a permission granted by the owner of an airport designated which allows the grantee to schedule a landing or departure at that airport during a specific time period.

While it’s kept mostly hush-hush, a leaked document revealed Oman Air set a record by paying £58 million for a pair of take-off and landing slots at Heathrow in early 2016. The Gulf carrier is understood to have bought the pair, which includes a highly prized early morning arrival, from Air France-KLM.

A year later, Scandinavian Airlines made news when it revealed that it sold two slot pairs at Heathrow for almost £60 million.

Pre-pandemic, European and airports instituted an 80/20 policy. That is; the airline allocated a slot must fly at least 80% of the allocated slots over a year. If you’re interested in the full details of the policy, here’s the full text of the EU Slot Regulation EEC 95/93.

There is good reason for this. It ensures the aviation market remains competitive and airlines are incentivised to trade or hand back unused slots so that other airlines can fly them instead, including new market entrants.

However, along came COVID-19 and an almost blanket ban on international travel. The 80:20 rules were suspended for six months in March 2020 to stop airlines running empty flights to hold on to their landing rights. This was extended for another six months in September 2020.

More recently, the UK government is operating a 70% rule throughout the summer of 2020, expecting passenger volumes to be somewhere similar to pre-pandemic levels.

So now these restrictions are back in place, at least in the UK, I wanted to take a look at the impact of so-called ghost flights over the past 2 years.


I recently came across a Freedom of Information request detailing the of the number of empty or almost empty flights that have left the UK by airport in each month since March 2020.

“Empty or almost empty” has been classified as flights with a load factor of 10% or less for this question. This means that the number of terminal passengers on the flight was no more than 10% of the number of available seats. Scheduled and chartered flights are included, as are flights to Oil Rigs.


Count of passenger flights operating from UK airports with load factor less than 10%

Count of passenger flights operating from UK airports with load factor less than 10%

Download graph.

You can see in March 2020, for most of which the 80:20 rules was in place, the number of ghost flights (1772) was at it highest during the pandemic, as airlines couldn’t fly passengers but were tied to the 80:20 rule before it was paused.

As travel started again in the Autumn of 2020, you can see the total number of ghost flights increasing again (1544 in September 2022, the second highest month for ghost flights). The assumption here is passengers simply didn’t want to travel but airlines ramped up the number of flights, thus load volumes were below 10% on operating flights.

In total there were 14,472 ghost flights over the period of Mar 2020 and Sept 2021.

Total ghost flights by airport Mar 20 – Sept 21

Reporting Airport Total (Mar 20 – Sept 21)
Kirkwall 0
Dundee 3
Exeter 10
Gatwick 1044
Manchester 1548
Heathrow 4910

View full table.

Kirkwall had no ghost flights, whereas Heathrow had 4910 over the period Mar 20 – Sept 21.

Though Kirkwall handled 14,247 flights in 2019. Heathrow handled 80,886,589.

Ghost flights Sept 20 – Sept 21 as % of normal flight volume (2019 (top 20 UK airports)

Ghost flights Sept 20 - Sept 21 as % of normal flight volume (2019)


Download graph.

Looking at the top 20 airports, Aberdeen has the largest percentage (0.019%) of ghost flights (Sept ’20 – ’21) versus 2019 (pre pandemic) total aircraft movements. One reason for this might be due to the types of flights. In the dataset I am using

Scheduled and chartered flights are included, as are flights to Oil Rigs.

Aberdeen serves oil rigs in the North Sea, and is therefore likely to operate transfers for workers which might usually have a load factor of less than 10%, even pre-pandemic.

Bristol (0.0047%) and Liverpool (John Lennon) (0.0046%) are the second and third respectively when it comes to worst offenders for total ghost flights as a percentage of normal operation.

Ghost flights estimated c02 emissions

Let’s assume a ghost flight operates on the same averages I used for my post on aircraft emissions for delegates travelling to last years COP 26 summit;

On average, passenger aviation emitted 90 grams of CO2 per passenger-kilometer in 2019

International Council on Clean Transportation

Ghost flights also fly planned routes (to ensure slots are kept at both airports). So let’s also assume the average flight distance used for calculations in the COP 26 post too; 3000 km (6000 km for both legs).

Estimated co2 emissions (tons) from UK ghost flights (Mar 20 - Sept 21)

Download graph.

Note, we’re assuming the average flight lengths to be identical across airports. In my calculation the major international airports cause the most emissions from ghost flights simply because the ran more of them.

The reality is that emissions will be much higher from these airports. This is because departing and arriving flights tend to fly much longer distances when compared to regional airports (where real emissions will likely be lower than my calculation for the same reason).

In total, I’ve estimated total emissions from ghost flights between March 2020 – Sept 2021 from listed UK airports was 7152.84 tons of CO2.

Ghost flights cost to the airline

Current jet fuel costs about $0.95 / litre.

According to this Wikipedia article, jet fuel weighs 0.81 kg/l.

A Boeing 787-9 burns 5.77 kg of fuel per km.

Therefore, per km, a 787-9 burns $6.77 worth of fuel per km ((5.77/0.81)*0.95).

Estimated fuel cost (USD) for UK ghost flights (Mar 20 - Sept 21)

Download graph.

Operating an airline is not cheap!

Assuming the average ghost flight flew 6000 kms, and the airline paid current costs $0.95/litre (total $40,620 per flight), airlines flying ghost flights out of Heathrow paid almost $200 million in fuel costs to operate these empty or almost empty flights.

In total I estimate over half a billion dollars ($538,052,520) was spent by airlines operating ghost flights from these UK airports during the pandemic.


This post only considers UK airports. Ghost flights are not unique to the UK. As noted previously, the 80:20 rules applies to US and EU airports too. Lufthansa has apparently operated over 18,000 ghost flights — if true, 4,000 more than all UK airport ghost flights combined.

It would therefore be a logical next step to compare ghost flights worldwide, and by airline (whilst also factoring in the improvements to my calculations as noted throughout this post).


I estimate over half a billion dollars ($538,052,520) was spent by airlines operating ghost flights from UK airports during the pandemic resulting in a total of 7152.84 tons of CO2 emissions.


  1. Data sources + data used in this post.

Ski Resort Economics: Billions Made from Lift Pass Sales Each Year

Skiing is an expensive sport.

I’ve talked previously about the price for a cup of coffee on the mountain.

Lift passes are another huge expense.

During busy periods, like school holidays, you can often be stood in lift queues with hundreds of other people.

It’s not long before you start thinking about just how much money all of these people have spent on their passes that day.

Back at my computer, sadly, I wanted to get a better idea of revenue resorts earn through the sale of lift passes.


Each year the International Report on Snow & Mountain Tourism ranks the most visited ski areas based on the average annual skier visits.

I ignored the 2019/2020 season due to COVID travel restrictions, and used the 2018/2019 season for the visitor numbers taken from the report. Visitor numbers are defined as unique daily lift tickets (e.g 1 person visiting over 6 days would count as 6 visitors).

Lift ticket prices were obtained for the current 2021/2022 season.


Total visitors numbers by top winter destination

Daily unique lift pass entries by resort country (2018/19)

Download graph.

France, Austrian and French resorts were home to over 51 million visitors each in the 2018/19 season. In comparison, the population of Austria is only 9 million (France is 67 million, US is 330 million).

Foreign visitor numbers by top winter destination

Foreign visitor numbers by top winter destination

Download graph.

Country Total visitors (19/20) Percent foreign visitors Total foreign visitors
Canada 18521000 12 2222520
United States 54250000 6 3255000
Switzerland 22274000 35 7795900
Italy 27603000 35 9661050
France 51009000 27 13772430
Austria 51700000 66 34122000

Full table.

Austria also received the highest number of total foreign visitors. 66% of the 51 million visitors (34.1 million) were foreign.

France received the second highest volume of winter ski tourists, 13.8 million.

Winter destination lift ticket income by country

For this section I assumed lift ticket costs from one of the major resorts in each country (see improvements), to work out potential lift ticket income across the entire country based on the visitor numbers.

Resort Country 6 day pass cost (EUR) 1 day pass cost (EUR)
Montgenèvre Italy 143.9 23.98
Les Portes du Soleil Switzerland 295 49.17
Val Thorens France 280 46.67
Whistler Blackcomb Canada 605 100.83
Kitzski Kitzbühel Austria 253 42.17
Jackson Hole United States 943 157.17

Full table.

North American resorts are much more expensive than their European counterparts — you could by a whole week (6 days) in Montgenèvre, Italy (144EUR) for one day in Jackson Hole, USA (157EUR).

Estimated total lift pass revenue (2018/19 visitors, 2021/22 lift ticket cost)

Download graph.

The visitor numbers combined with lift prices means the USA lift ticket revenue alone could have been as high as 8.5 billion EUR. Even assuming an average of half this cost for lift tickets, that’s still 4.25 billion EUR.

French and Austrian resort in comparison could have earned between 2.2 and 2.4 billion EUR, and even a conservative estimate would probably see these resorts pulling in over 1 billion from lift ticket sales.

Coffee sales income by country

As linked earlier, my post about coffee prices on the mountain found the average price for a cup being £2.91 (3.51 EUR) (in 2018).

Estimate coffee sales revenue in winter resorts (2018/19)

Download graph.

Assuming each visitor (one day lift ticket) bought one coffee on that day, the resorts in France, Austria and the United States could have each earned hundreds of millions of EUR in coffee sales alone. When you start adding in meals, these numbers get significantly bigger (the average cost of a lunch at a restaurant was 20.11 GBP (24.25 EUR), per person.


This post contains very broad estimates. I assumed lift prices for some of the most popular resorts in each country, and thus likely some of the most expensive. Doing this analysis on a resort by resort basis would make for a very interesting comparison.


Lift tickets are worth billions of EUR in revenue to the major snow resort countries. In the USA, it could be as high as 8.5 billion EUR from lift ticket sales alone during winter.


  1. Data sources + data used in this post.

The Worlds Longest Passenger Train Journeys

I’ve sat on many a long train journey in the UK, not necessarily due to the distance travelled, more so the speed of travel.

For a long time traversing Russia on the trans-Siberian railway has been on my bucket list. I assumed this to be the longest train journey in the world.

However, recently I read a post titled “The Longest Train Ride in the World“, documenting the longest possible trip when taking connections at stations.

It mentioned:

“The longest single uninterrupted train journey, including transfers, does indeed stretch beyond London and Beijing.”

Living in London I’d never heard of this trip, so I decided to do some digging.


In this post I’ve only considered trips where you do not have to connect to other services.

It turns out that the London to Beijing passenger route is a mix of different routes; starting off with the Eurostar, then a train to Russia, and finally for the longest part of the journey, the trans-Siberian route. Thus, it is disqualified from this analysis.

That said, there is a direct freight route from London to Yiwu, China. More on that in next months post.


Longest Routes by distance

10 longest passenger rail journeys (2021)

Download graph.

Rank (length km) Origin – Destination No. of Stops Distance (km) Ave Distance between stops (km) Scheduled running time (hours)
1 Moscow (Russia)-Pyongyang (North Korea) 157 10267 65.39 206
2 Moscow (Russia)-Vladivostok (Russia) 73 9259 126.84 167
3 Moscow (Russia)-Beijing (China) 33 8984 272.24 145
4 Moscow (Russia)-Beijing (China) 33 7826 237.15 127
5 Kislovodsk (Russia)-Tynda (Russia) 107 7734 72.28 143
5 Moscow (Russia)-Neryungri (Russia) 91 6950 76.37 140
7 Guangzhou (China)-Lhasa (China) 12 4980 415.00 52.1
10 Shanghai (China)-Yining (China) 25 4742 189.68 49.2
11 Shenzhen (China)-Urumqi (China) 21 4666 222.19 49
12 Guangzhou (China)-Urumqi (China) 28 4663 166.54 48
13 Changchun (China)-Sanya (China) 23 4647 202.04 53.3
14 Nanning (China)-Urumqi (China) 31 4617 148.94 59.6
15 Harbin (China)-Kunming (China) 49 4574 93.35 63.3
16 Changchun (China)-Urumqi (China) 33 4507 136.58 56.8
17 Toronto (Canada)-Vancouver (Canada) 66 4466 67.67 92
18 Chicago (United States)-Los Angeles (United States) 40 4390 109.75 65.3
19 Shanghai (China)-Lhasa (China) 12 4373 364.42 46.3
20 Urumqi (China)-Fuzhou (China) 23 4334 188.43 50.2

Full table.

Moscow to Pyongyang, traversing Siberia (and the trans-Siberian railway) is the longest journey at 10,367km and taking 206 hours, or 8.5 days (206/24).

The 6 longest journeys by distance all start or finish in Russia.

Longest Routes by speed

Ave speed of 20 longest passenger rail journeys (2021)

Download graph.

Origin – Destination Ave. speed (km/h)
Moscow (Russia)-Pyongyang (North Korea) 49.84
Moscow (Russia)-Vladivostok (Russia) 55.44
Moscow (Russia)-Beijing (China) 61.96
Moscow (Russia)-Beijing (China) 61.62
Kislovodsk (Russia)-Tynda (Russia) 54.08
Moscow (Russia)-Neryungri (Russia) 49.64
Guangzhou (China)-Lhasa (China) 95.59
Shanghai (China)-Yining (China) 96.38
Shenzhen (China)-Urumqi (China) 95.22
Guangzhou (China)-Urumqi (China) 97.15
Changchun (China)-Sanya (China) 87.19
Nanning (China)-Urumqi (China) 77.47
Harbin (China)-Kunming (China) 72.26
Changchun (China)-Urumqi (China) 79.35
Toronto (Canada)-Vancouver (Canada) 48.54
Chicago (United States)-Los Angeles (United States) 67.23
Shanghai (China)-Lhasa (China) 94.45
Urumqi (China)-Fuzhou (China) 86.33

Full table.

China holds 12 of top 20 longest routes, though it’s clear China’s railways are leading the way in terms of speed of travel.

The fastest of these is Guangzhou (China)-Urumqi (China) where the average speed is 97.15km/h! If it were not for Chicago (United States)-Los Angeles (United States) (11th fastest at ave speed 67.23 km/h), China would hold the top 12 positions in terms of speed.

If the longest passenger route, Moscow (Russia)-Pyongyang (North Korea) which currently takes 8.5 days at 49.84 km/h, travelled at a similar speed, it could do the journey in a little over 4 days.

Longest Routes by cost

For this, I used the earliest possible date of departure in March 2022 (search performed in December 2021) for each route. Note, the figures you see below are the ones I could obtain. Many ticket fares proved difficult to determine due to translation issues.

Each service offered 3 classes, the average of which you see below. However it is important to note, third class in Russia might not be the same as third class in China, and so on.

Rank (length km) Distance (km) Origin – Destination Ave cost (USD)
2 9259 Moscow (Russia)-Vladivostok (Russia) 998.44
7 4980 Guangzhou (China)-Lhasa (China) 145.77
17 4466 Toronto (Canada)-Vancouver (Canada) 1,784.00
18 4390 Chicago (United States)-Los Angeles (United States) 395.33

Full table.

The prices are quite telling of the journeys. Moscow (Russia)-Vladivostok (Russia) and Toronto (Canada)-Vancouver (Canada) are touted to tourists, perhaps an explanation of the higher prices.

Though the Moscow (Russia)-Vladivostok (Russia) mean average is somewhat skewed — a cheap 3rd class fair can be bought for $387.45 (vs. $1,602.40 for 1st class).


Ticket prices are surprisingly hard to come by on many of these routes (a large part down to my lack of Chinese and Russian language skills). Personally, I’d really like to do a more accurate analysis of train ticket prices around the world.


The longest passenger train journeys in the world all start/finish in Russia, the longest of which being Moscow (Russia)-Pyongyang (North Korea) at 10,267 km (or 8.5 days).


  1. Data sources + data used in this post.

Airports Had to Deal With Over 2 Billion Litres of Airplane Toilet Waste in 2019

We already know airplanes don’t have enough toilets.

But where does all the waste go once the vacuum has sucked the waste out of the toilet bowl?

No, it doesn’t get ejected.

But on a busy flight recently (Easyjet, many drunk tourists, lots of trips to the toilet) I started to think; how big is the waste tank? Can it ever get full?

Plane waste tank sizes

Manufacturer Model Total waste capacity (L)
Boeing 747-400 1135
Boeing 787-8 1628
Airbus A380 2096

Using publicly available data sheets for aircraft, I found the above specifications for each waste tank.

Waste capacity per passenger

Let’s assume the toilets are exclusively for urination to simplify the calculations.

According to Healthline:

The normal range of urine output is 800 to 2,000 milliliters per day if you have a normal fluid intake of about 2 liters per day.

Let’s take also assume this is passed during waking hours, so 16 hours (960 mins), and that people are awake on flights.

At the lower estimate (800ml) that’s 50ml per hour. At the higher estimate (2000ml) that’s 125ml per hour.

The planes for which I have collected statistics for all fly predominantly long haul routes. Let’s assume that to be 6 to 11 hours.

Average human urination rates (ave volume per hour)

Download graph.

Hour Low (urinate vol ave ml) High (urinate vol ave ml)
1 50 125
2 100 250
3 150 375
4 200 500
5 250 625
6 300 750
7 350 875
8 400 1000
9 450 1125
10 500 1250
11 550 1375
12 600 1500

Full table.

Assuming a full 11 hour flight, best case, each passenger will pass 550 ml, worst case, 1375 ml.

Manufacturer Model Total waste capacity (L) Total pax (typical) Total wast capacity per pax (L)
Boeing 747-400 1135 416 2.73
Boeing 787-8 1628 467 3.49
Airbus A380 2096 525 3.99

Even on the 747-800 which has the lowest waste tank volume per passenger, this still leaves room for 1 extra litre in the tank per passenger. On the A380, there’s over 2.5 spare litres of waste tank capacity per passenger.

It’s unlikely they’ll overflow, even on a particularly boozy flight.

Dealing with waste

When a plane lands, most of you would have noticed a fuel truck and the baggage carts arrive to empty the plane and prepare it for the next flight. You might not have seen the waste disposal truck arrive to empty the tank (generally at the back of the plane).

Let’s assume the average flight time for planes arriving at airports is 4 hours (so planes are carrying between 200ml – 500ml of urine per passenger).

Busiest airport by pax Apr 2021 graph

Download graph.

Airport Country Passengers Apr 2021 (mil) Total waste low Apr 2021 (urinate vol ave l) Total waste high Apr 2021 (urinate vol ave l)
Guangzhou (CAN) China 2.572 5144000 1286000
Atlanta (ATL) United States 2.459 4918000 1229500
Dallas (DFW) United States 2.364 4728000 1182000
Chengdu (CTU) China 2.329 4658000 1164500
Denver (DEN) United States 2.182 4364000 1091000
Beijing (PEK) China 2.061 4122000 1030500
Shenzhen (SZX) China 2.055 4110000 1027500
Shanghai (PVG) China 1.994 3988000 997000
Charlotte (CLT) United States 1.889 3778000 944500
Shanghai (SHA) China 1.863 3726000 931500
Chicago (ORD) United States 1.768 3536000 884000
Hangzhou (HGH) China 1.669 3338000 834500
Orlando (MCO) United States 1.637 3274000 818500
Phoenix (PHX) United States 1.585 3170000 792500
Los Angeles (LAX) United States 1.537 3074000 768500

Full table.

Let’s assume the upper end of the estimates, the world’s busiest airport by passenger volume, Guangzhou, would have handled 5.14 million litres of pee from 2.57 million passengers in April 2021. At that rate, they’d be dealing with over 60 million litres of pee every year.

For reference, an Olympic sized swimming pool holds 2.5 million litres.

According to ICAO:

the total number of passengers carried on scheduled services rose to 4.5 billion in 2019

Assuming that low estimate (200ml per pax), that equates to 900 million litres of waste from all planes in 2019.

If we consider each passenger passed an average of 500ml, that’s 2.2 billion litres (or 880 Olympic swimming pools… which to me doesn’t sound that much!).


The world’s busiest airport by passenger volume, Guangzhou, could have handled 5.14 million litres of pee in April 2021. On a worldwide scale, airplane yearly toilet waste is estimated to be 900 million – 2.2 billion litres total. Despite this volume of pee, it’s unlikely the waste tank on a plane will ever reach capacity.


  1. Data sources + data used in this post.

Santa’s Fulfilment Center Handles Over £162 Billion of Gifts in Single Day

It’s nearly Christmas!

Time for a festive post.

No expensive Christmas trees or private jets this year — I’d like to keep the number of people going into debt to fund Christmas down.

This year we’ll rely solely on Santa.

I’ve got used to Amazon Prime delivery, but even Amazon do not have the delivery capacity of Santa… yet.

Let’s start by looking at how many gifts Santa will be delivering this year…

Estimated number of households by continent

Download chart.

Examining total households (excluding persons living inside collective living quarters, such as hotels, rooming houses and other lodging houses, institutions and camps) there are 2,174,034,795 households.

There are 86,400 seconds in a day. So Santa needs to deliver to one household every 0.00003974177 seconds (86,400/2,174,034,795). No time for a snack break!

Lets for a moment ignore continents where the Christianity is less dominant (namely Asia), we have about 1 billion households. In this case Santa needs to deliver to one household every 0.0000864 seconds (86,400/1,000,000,000). Still no time for a snack break!

According to the UN, the world population in 2021 is 7.9 billion.

If we consider the official Premier League match ball, the Nike Premier League Flight which costs £125 at retail, Santa’s total bill for presents would be £987,500,000,000 (125*7,900,000,000) — almost £1 billion! Even if he chooses a relatively cheap £20 ball (or is able to negotiate a bulk discount), we’re still talking in the 100’s of billions of Pounds, £158,000,000,000 (20*7,900,000,000)

It’s hard to visualise that many people. Although I’m informed by National Geographic:

Standing shoulder-to-shoulder, the entire world’s population could fit within the 500 square miles (1,300 square kilometers) of Los Angeles.

But what do 7.9 billion presents look like?

In the interests of fairness, let’s assume each present is the same. This year everyone is getting a football (soccer ball). This makes it easier for Santa as it can be delivered in a cube shaped box.

Rules state that a size 5 ball must be 68 to 70 cm in circumference. So the maximum radius is 35cm. Let’s assume the box’s dimensions are 35cmx35cmx35cm (a cube).

Laid side by side the boxes would cover 2,765,000km (0.00035km*7,900,000,000). Placed around the equator (40,075km), the presents could be laid around the planet almost 69 times (2,765,000/40,075).

The earths surface area is 510.1 million km². One side of each box has a surface area of 0.00035km² and all boxes together, 2,765,000km². So all presents, each laid on the ground, the would cover 1/184 (510,100,000km/ 2,765,000km) of the earths surface.

ESO estimate the Pacific Ocean holds 707.5 million km3 of water. The presents, have a total volume of 2,765,000km3 (0.00035km3*7,900,000,000). They would 1/256 of the Pacific Ocean (707,500,000/2,765,000)!

Amazon’s sortable fulfillment centers are around 800,000 square feet in size (74,322 m2). Let’s assume that’s 743.22m (w) x 100m (l) x 30m (h) = 2,229,660m3 (or 0.00222966km3)… meaning Santa is going to need 6,166 of them.

An official size 5 football weighs between 410 – 450 g without packaging. Therefore accounting for the upper end of this allowance to cover potential packaging weight, the total weight of all the balls would be 3,555,000,000 kg (0.450*7,900,000,000) or 3.5 million tons.

The A380-800F (the Airbus A380 freighter) has a payload capacity of 150,000kg. So if the sleigh breaks down, Santa would need a minimum of 23,700 cargo flights (3,555,000,000kg/150,000kg). According to one source only 27 of these freighters have been produced, so would need 878 flights by each plane to distribute the presents.

Of course, the space vs weight ratio is an important factor. The A380-800F has a total hold volume of 1,134m³ (or 0.000001134km3). Yes, Santa would need A LOT more planes!

To put that into perspective, National Geographic estimate:

The world generates at least 3.5 million tons of plastic and other solid waste a day


Now, Santa can’t send these presents unwrapped.

A cursory glance at Amazon shows me a 4 pack of 500cmx70cm Christmas wrapping paper for about £10. Put another way £10 buys you 0.02kmx0.0007km (a total surface area of 0.000014km2).

Each of our presents has a surface area of 7.35e-7km (0.735 m2) (6*0.00035km*0.00035km).

So one pack of wrapping paper is enough for 19.048 presents 0.000014/(7.35e-7), wrapped perfectly with no overlap and no mistakes.

But we need enough for 7,900,000,000 presents!

In total, that’s 414,741,706 (7,900,000,000/19.048) packs. At £10, ignoring any bulk discounts, that’s £4,147,417,060 (£4.1 billion) just to wrap the presents!

Thinking about the environmental cost for a minute…

Thin and straight trees are used for paper making and the average log of such a tree would weigh about 50 kgs.

Roughly about 50% of the wood is converted into wood pulp. So 50 kgs of wood would make 25 kgs of pulp. Roughly, 1kg of pulp makes 1.2 kg of finished paper. Hence, 25 kgs of pulp would make 30 kgs of paper.

Assuming a 75 GSM (grams per square meter) paper weight, our pack of paper weighs 1050g (20mx0.7m*75g) or 1.05kg or 0.00105 tonnes.

The total weight of all paper required would be about 435,479 tonnes (0.00105t * 414,741,706), requiring 8,709,580 trees (435,479/0.05).

For a frame of reference, Nature report there are about 3.04 Trillion trees on Earth. According to

Throughout the world, about 900 million trees are cut down annually. This equates to about 2.47 million trees cut down every day.

So it would take 4 days worth of Earths tree-felling activities to supply the wood needed for Santa’s wrapping paper needs.

All that’s left now is tape. Let’s assume perfection, that is; we only need tape for one face (shout out to those who can neatly wrap presents, a skill I don’t posses). That’s 140cm of tape per box (35*4).

In total that’s 11,060,000km of tape needed (0.00140*7,900,000,000) — almost 276 times (11,060,000km/40,075km) around the world.

Quickly looking at Amazon, I can get 50m of Selotape for £1. Santa therefore needs 221,200,000km rolls of tape (11,060,000km/0.05km) at a total cost of £221,200,000.

I think we’ll save Santa Claus the need to write cards for each present this year…

  • Presents: £158,000,000,000
  • Wrapping Paper: £4,147,417,060
  • Tape: £221,200,000
  • Total cost (exc. Reindeer food): £162,368,617,060


Have a great Christmas for those celebrating!


  1. Data sources + data used in this post.

An Estimated 11,000 International Delegates at COP 26 Will Generate 5400 Tons of CO2 emissions

At this years G7, it was only a matter of time before news stories relayed the climate outrage of politicians and their entourages flying (sometime only 100’s of km’s) into Cornwall, UK.

I share the outrage at such stories.

I expect the same at this years COP 26 conference in Glasgow this year.

According to the Energy and Climate Intel Unit:

It will involve upwards of 30,000 people in the city, representing over 200 countries, businesses, NGOs, faith groups and many more.

With such a large number of delegates in attendance, the climate footprint will be significantly greater than the G7. As my mind started whirring, I wondered if it was possible to quantify just how much impact to the environment travel to and from the conference could cause.


In 2019, the UN published the list of all delegates attending the COP 26 Summit.

In summary;

States/organisations Participants
Parties 196 11406
Observer States 1 8
United Nations Secretariat units and bodies 28 306
Specialized agencies and related organizations 23 400
Intergovernmental organizations 76 652
Non-governmental organizations 1049 7417
Media 844 2165
2217 22354

Download full dataset.

I’ll make a very rough assumption that half the attendees will come from the UK, so let’s round down and say 11,000 attendees will come from outside the UK.


CO2 per passenger

Now we know 11,000 will be flying into Glasgow, lets calculate average fuel consumption.

One way to calculate CO2 emissions is from fuel consumption per flight. Using old data (published 2010) comparing a Boeing 737-400 (short haul) and Boeing 747-400 (long haul) and average load factors (note passenger load factors will always vary);

Boeing 737-400 Boeing 747-400
Distance (kms) 926 5556
Fuel used (tons) 3.61 59.6
Seats 164 416
Ave load factor % 65 80
Ave pax 107 333
Fuel use (g) per pax km 36.57118312 32.23299828
CO2 emissions (g) from aviation fuel per gram of fuel 3.15 3.15
Total CO2 emissions (g) per pax km 115.1992268 101.5339446

OK, so both of these planes are quite old (both have many newer versions). Fair point. I used them as the above statistics were freely available.

According to

The aviation sector’s short-term goal to improve fleet fuel efficiency by an average of 1.5% per annum from 2009-2020 is on track, with current analysis showing a 2.3% improvement on a rolling average − an efficiency improvement of 17.3% since 2009.

Let’s assume then that newer planes are 17.3% more efficient, we get an emissions figure of 89.316 g CO2 per passenger km ((101+115)/2)*(1-0.173).

This figure is very close to on reported from the International Council on Clean Transportation:

On average, passenger aviation emitted 90 grams of CO2 per passenger-kilometer in 2019

Both calculations will be a little optimistic for COP 26, as the original calculations were based on high load factors, which have reduced significantly, and still remain so, because of COVID. But let’s go with it for now…

Average travel distance per passenger

Attendees travel from all over the world, but it would appear from reports most travel will be from European delegates given location. I wasn’t able to easily break the 2019 delegate list by country (although that was held in Chile, so the delegate count by country will no doubt be quite different this year).

There has already been coverage that some nations not being able to field delegates this year due to travel restrictions. Namely in Asia, Oceanic, and African countries.

Given the above, I considered the edge of Europe, Istanbul as a rough guess for “average distance” attendees will travel.

Let’s assume then, that on average, delegate travel the distance of Istanbul to Glasgow (shortest distance = 2912 km, rounded up to 3000km, and doubled for the return journey 6000km).

Delegate emissions from flights

We now estimate delegates will have a round trip to Glasgow of about 6000km. Therefore, one passenger will produce 540kg CO2 pin CO2 emissions flying to and from the conference (0.09*6000).

For all 11,000 delegates, that’s 5,400,000 kilograms of CO2 or roughly 5400 tons of aviation CO2 emissions to fly international delegates in and out of Glasgow.

Note, this does not factor in the increased warming effect other, non-CO2, emissions, such as nitrogen oxides, have when they are released at high altitudes can also make a significant difference to emissions calculations.

What if…

Due to Glasgows position in the UK, and the fact that the UK is an island, it’s likely most people will fly into to Scotland directly.

Perhaps some Europeans will take advantage of the Eurostar to London, and then a 5 hour train journey to Glasgow, but my assumption is that a very small number of attendees will choose this route owing to the low cost, much faster air routes.

However, let’s compare aviation to other available methods of transport.

Total C02 emissions (tons) for 11000 international delegates at COP 26 by transport type

Download chart.

According to DEFRA research, driving alone is almost twice as bad as flying (11,286 tons of CO2 emissions if all international delegates drove alone)!

However, remember, we have ignored +50% of attendees, assuming them to be UK residents… many of whom will be very likely to drive (probably alone)! The 11286 tons represents a 6000km round trip. Dividing this by 10 (600km round trip for UK residents), still leaves 1128.6 tons of CO2 emissions for the driving domestic delegates.

Car sharing (assuming 4 delegates) almost cuts emissions in half from driving alone. Though, roughly, two people to a car (probably more realistic) delivers the same emissions as a comparable plane journey, per passenger, over the same distance.

Unsurprisingly rail is much more efficient when it comes to CO2 emissions per passenger (lets hope most UK delegates choose this over a solo car car journey as rail is 6.3 times more efficient). Sadly, the UK doesn’t have a highly efficient rail infrastructure like that of the Eurostar (which is over 4 times more efficient than general rail). Is HS2 still happening?


This really is a “back-of-the-napkin” analysis. Journalists, please don’t credibly cite these figures unless you make this clear!


The 22,000+ delegates (under-estimate) travelling to the COP 26 conference in Glasgow are likely to generate over 10,000 tons in CO2 emissions.


  1. Data sources + data used in this post.

There are over 200 electrically propelled aircraft projects in development (though most have a range far below 200 kilometers)

269 billion litres of jet fuel was burned in 2017 — Enough fuel to fill 5.4 billion VW Golfs

We’ve all seen the headlines; commercial aviation is bad for the environment.

If the industry bounces back to pre-COVID growth, which most analysts believe will be the case, an alternative to fossil fuels is desperately required to stop the devastating impacts of global warming.

Whilst electric cars are now at a point where they can compete (and out-do) oil powered cars, the size and power requirements of planes have meant further improvements to battery technologies are required before we see electric aircraft.

According to Elon Musk in 2019:

But two years is a long time in battery technology.

How close are we to electric powered commercial aviation? I decided to take a look at the current state of the market.


For this analysis I used a report produced by Roland Berger in 2020 on the electric aircraft industry which reported detailed information about 218 planes. The numbers include planes across all stages of development from concepts to airworthy models (albeit there are currently very few).


Count of Aircraft by Market (2020)

Count of Aircraft by Market

Download chart.

Download full table.

Urban air taxis still dominate the scene, representing about 45% of all aircraft, though general aviation (typically recreational planes) is close behind, comprising 85 projects globally.

Count of Aircraft by Type (2020)

Count of Aircraft by Type

Download chart.

Download full table.

48% of all projects are VTOL (vertical take off and landing) craft. Given most projects are focused on short range / intercity aviation, likely operating in cities, VTOL makes most sense.

Count of Aircraft by Power Type

Count of Aircraft by Power Type

Download chart.

Download full table.

Battery power (including hybrid battery / oil) powered planes account for 97% of all projects.

Solar and hydrogen make up the rest. Solar being employed in concepts for general aviation for small recreational planes. Hydrogen seems to be attractive to manufacturers building autonomous craft, like HES developing the Element One.

The majority of all these planes are using propellors for propulsion although a tiny number of projects like Airbus’s E-Fan-X are using turbofan engines.

Pax capacity of battery powered aircraft (2020)

Pax capacity of battery powered aircraft (2020)

Download chart.

Download full table.

Looking at the 122 battery powered plane projects that report passenger capacity, the maximum capacity is 19, touted by Heart Aerospace’s concept.

The mean/median passenger load is 2, which highlights how most planned aircraft are aimed at consumers (potential replacements for cars).

Range of battery powered aircraft (inc. estimates) (2020)

Range of battery powered aircraft (inc. estimates) (2020)

Download chart.

Download full table.

Looking at the 53 battery powered planes that have reported range in distance (keep in mind these are for the most part estimates), all but 4 have ranges below 1000km. The median advertised range of these aircraft is 150km (mean is 298km).

The aircraft with the longest range, 2000km, is Avioneo’s conceptual 2345 craft. Wether it makes it to market is questionable, but it currently boasts costs of 0.12EUR per/km with a cruise speed of up to 300 km/h.


Many of these projects are small and will probably never enter the market. Most information being reported is estimated (and in many cases not even estimates, e.g. for range, are provided), as the planes are still mostly in conceptual phase.

At best this post can be considered a projection on the market. It needs revisiting in one years time, when it is likely a good number of viable planes will be entering early construction and some certifications being conducted.


The median advertised range of battery powered aircraft is 150km… most are still in concept phase. Electrically powered international commercial aviation is still some way off. On the other-hand, city based short-haul aviation is looking like a brand new market that might be realised in the near-future.


  1. Data sources + data used in this post.

The true scale of COVID travel restrictions on tourism based economies (72% of Macau’s GDP comes from tourism)

In 2017, or 3 BC (Before COVID), I looked at the amount the average tourist spent in a country compared to the average incomes of the populations.

Fast forward to 2019, or 1 AC (After COVID), and the international travel industry is still full of restrictions for travellers having a significant impact on tourism.

Listening to stories coming out resorts here in Europe that are built around tourism, it is heart-breaking to hear of the challenges they are facing after another year of reduced tourist numbers.

I can imagine many countries entire economies are driven from tourism income. In this months post I decided to take a look the countries most reliant on tourists, and thus likely most impacted by another summer of travel restrictions.


The World Travel and Tourism Council Data have curated an annual dataset, starting in 2000 to 2019, that details the tourism percentage that accounts for each countries GDP. Tourism considers both domestic and international visitors.

For actual GDP figures, I used data produced by the World Bank.


Tourism as % of countries GDP (2019)

Tourism as % of countries GDP (2019)

Download chart.

Rank by GDP/tourism (2019) Country % tourism of GDP 2019
1 Macau 72.00
2 Maldives 66.10
3 Seychelles 65.80
4 St Kitts and Nevis 62.60
5 Grenada 55.80
171 Moldova 3.20
172 Libya 3.10
173 South Korea 2.80
174 Democratic Republic of Congo 1.80
175 Papua New Guinea 1.60

Download full table.

In Macau 72% of its GDP come from tourism ($39.7 billion). Not surprising given much of their economy is driven by Chinese tourists coming to gamble in its numerous casinos (so with more relaxed restrictions, probably not as affected as Las Vegas). Unsurprisingly, tropical Indian and Caribbean islands are all heavily reliant on tourism (and likely much more impacted by travel bans).

South Korea is one of the least reliant countries on tourism. Just 2.8% of their GDP came from tourism in 2019! Other developed countries have single digit percentage reliance on tourism (e.g. USA 7.8%, France 9.6%), but even 1% increase is a significant amount of money when you’re talking about GDP’s in the trillions of dollars (US GDP was 21.4 trillion in 2019).

Change in tourism as % GDP (2000 – 2019)

Change in tourism as % GDP (2000 - 2019)

Download chart.

Rank by GDP/tourism (2019) Country Tourism % of GDP 2000 – 2019 change
7 Cape Verde 28.90
1 Macau 24.70
10 Belize 22.20
2 Maldives 21.30
4 St Kitts and Nevis 18.50
8 St Vincent and the Grenadines -6.50
9 Antigua and Barbuda -8.00
25 Cyprus -8.40
6 Vanuatu -11.10
52 Bahrain -16.20

Download full table.

Many economies have tried to boost incomes through tourism in the last few decades. Macau reliance’s on tourism was 24.7% higher in 2019 than in 2000.

Other countries have suffered in a reduction in tourists due to various factors (political, environmental and economic), or have realised a dependance on tourism alone can be risky… as highlighted by COVID-19 travel restrictions.

USD value of tourism vs tourism as % of GDP (2019)

USD value of tourism vs tourism as % of GDP (2019)

Download chart.

Rank by GDP/tourism (2019) Country Tourism % of GDP 2019 Tourism value 2019 (USD)
113 United States 7.80 $1,671,791,526,366.00
71 China 10.90 $1,556,513,183,949.98
120 Japan 7.50 $379,865,465,670.34
101 Germany 8.60 $332,056,625,991.37
72 United Kingdom 10.90 $308,558,672,344.41
162 Burundi 4.30 $129,529,284.66
20 Sao Tome and Principe 27.40 $117,114,460.87
82 Comoros 9.90 $115,418,150.08
30 Tonga 20.60 $105,544,112.24
34 Kiribati 19.00 $36,982,968.37

Download chart.

$1.67 trillion dollars of the US economy (7.8% of GDP) was earned through tourism in 2019. This figure includes domestic and international tourism. With many international citizens still banned from entering the US, it is very likely 100’s of billions of Dollars have been lost in tourism receipts over the last 18 months.

The $32.9 million Kiribati earned through tourism (19% of GDP) might seem very small (some football players earn more than that in a year), but it’s worth noting the actual population of the country in 2019 was only 117,606.


Tourism is a very broad term that includes hotel bookings, travel, attractions, and much more. It would be valuable to know how tourism money is spent in each country, and as a result, which industries in each country suffer most from a reduction in tourism.


72% of Macau’s GDP comes from tourism, that’s $39.7 billion… and that number is growing (by 24.7% between 2000 and 2019).


  1. Data sources + data used in this post.