Russian Tank Problem

 If one Russian tank storage base runs out how many tanks are left in the rest? I saw a video today about how the 1295th  tank storage area in Russia is empty.

---

There is a famous estimation problem called the german tank problem. If you know tanks have a serial number on them from 1 up as each new one is produced. And you find some broken tanks with serial numbers on them. Can you estimate how many tanks N have been created?

---

But I came across a related problem today. If you have some tank storage bases and one of them runs out of tanks what can you guess about the total number of tanks left?

This involves some assumptions. Each tank storage base draws at the same proportional rate. As in if one base has 100 and another has 100 each will have a 50:50 chance of having the next tank removed. And in the case where the second has 1000 tanks it will have a ten times higher chance of the next removal.

Bases should have a similar proportion of fixing up sheds and staff per tank so I think this is a reasonable guess. Unless some base you know to be much more likely to have high repair speeds and rates.

So this tank storage base is empty according to this short video 1295th - Empty of Tanks in 2024 December Satellite Imagery. Looking at the video it does not seem to be high tech refurbishing place. And if it was other tanks to repair would be moved there to also get repaired so it would not be empty first. it does seem small only having 325 tanks 2 years ago. But with the proportional withdrawal assumption that doesn't matter hugely.

This is the code to simulate how many tanks are left in all bases once one runs out. 

Assuming random withdrawals are random.

There is less tanks left than I expected. Once one store is empty less than 10% of the tanks remain. 

And this is with various numbers of bases and initial number of tanks in the base.

Ireland's Electricity Simulation

What will Ireland's Electricity mix look like in 2030 assuming

1. Demand will increase by the amount Eirgrid predict
2. 2030 will be an hour by hour weather copy of 2023.*
3. We will build the wind turbines we say we will.
4. Solar and battery trends keep going as they have for over a decade
5. The grid can handle the new power sources + demands

The code for this projection is here. 
* obviously the weather in any hour of the year won't be the same as that hour in 2023. But by copying a year you get to see how wind and solar peak and lull in an accurate way. Combining more years into an average loses the 'what happens if its dark and calm for a week' accuracy of copying one years weather.

At the moment Ireland gets a lot of wind power. But all the white here under the black line is met by fossil fuels and imports. How much white under the demand line will there be in future?

As wind, solar and battery roll out the amount of fossil fuels and imports needed decreases.

To take one example with these trends continuing in June 2030 Wind and Solar cover a lot of the power needs. With Battery holding enough to smooth out gaps 

On a day view of the year you can see solar has started to cover a lot of low wind times.

and by 2033 continuing trend growth of Solar really would start to cover a lot of demand

 

year	demand_twh	solar_twh	wind_twh	waste_twh	battery_used_twh	unmet_twh
2023	39.7	1.0	13.7	0.0	0.0	24.9
2024	41.2	1.4	16.2	0.0	0.0	23.7
2025	42.9	1.9	19.1	0.2	0.1	22.0
2026	44.6	2.6	22.6	0.7	0.2	20.0
2027	46.4	3.6	26.6	2.0	0.3	17.9
2028	48.2	4.9	31.4	4.3	0.6	15.7
2029	50.2	6.8	37.1	7.9	1.0	13.3
2030	52.2	9.4	43.7	13.3	1.7	10.6
2031	54.3	13.0	51.6	21.1	2.9	7.9
2032	56.4	17.9	60.9	31.8	4.3	5.1
2033	58.7	24.7	71.8	46.1	5.3	3.0
2034	61.1	34.1	84.8	65.1	5.4	1.9
2035	63.5	47.1	100.0	90.0	5.3	1.2
2036	66.0	64.9	118.0	122.6	5.2	0.5
2037	68.7	89.6	139.3	165.1	4.8	0.1

Predictions of ten plus years are really not to be trusted. But the reduction to 10.6Twh in 2030 from 25Twh is something this government we just elected controls. We make planning and other decisions all the time that decide if we meet these trend increases in demand, solar, wind and batteries. We can decarbonise fast enough to prevent billions in fines in 2030 if we follow this path.