Here are my 2 cents.
I considered OP's scenario, but wanted to add a twist to it.
I considered the Vampire does not have to be awake all time and could use some nights of rest if he or she finds a hiding spot that will last for a while. Thus, here is my typology of situations he or she might face, each night:
- CASE A - Night of rest (no blood points spent)
- CASE B - Has to awake to do some chores and incidents make him use
disciplines (1 BP for the awakening and between 1 and 2 extra ones)
- CASE C - He is awakened by imminent danger, which makes rises the BP
The overall expense is between 1 and 3 BP each night, as posited by OP. I biased my simulation towards Case C, which is a 50% random event (e.g. Werewolves storming your hideout) and then, out of the remainder 50%, I sorted randomly the other two options.
Note: The BP expense is always between 1 and 3. I could have made it easier, but this Case-driven simulation is the result of a previous arrangement in which I conceived of the possibility of spending 0 BP on some nights of rests, which is clearly denied by @DannnyCuppen, who kindly pointed out the fact p. 268 of the V20 edition says 1 BP is spent daily during slumber, and even twice is there is a daily awakening followed by another session of slumber.
Thus, I believe my simulations should account for the high-tension scenario put forward by OP, while, at the same time, accounting for occasional days of rest or basic chores on the remaining days.
For 100 simulations run over a period of 2 years, the BP expenditure ranges around 1800 BPs... That is, around 900 BP per year.
After doing this, I considered the following:
- Taking 1 BP from a human equals to about 1 donation
- Although blood volume is recovered in 24-48 hrs, haemoglobin takes 14 days on average to be recovered
- Suggested recovery times are 112 days (i.e. 16 weeks -- click here for a UK reference), but may be as short as 56 days for men and 84 days for women (see what WHO says at page 44 -- paragraph 4.6.2)
- Over 2 years of donation, donors who stick to the shortest time frame might develop iron deficiency... Hence, I freely assumed each donor who is subject to feeding events at a certain ratio must be given a 2 years break every 2 years of donor activity
Another assumption is that each BP requires a feeding event from a separate individual, since OP did not mention the need for our Vampire to cover up his activities.
In my next simulation, I considered how many BPs are needed each day, with a "refill ASAP" logic in mind.
I also accounted for the fact that "donors" become available again after 84 days (men) and 56 days (women) (see above). I also randomised male and female feeding events, so that, for every feeding event, the re-entry point in my feeding pool could vary.
I ran this crude simulation (very crude, actually) 54 times. The resulting feeding pool never exceeded 203 individuals and never went below 185 individuals.
Now, considering my 4-years refresh rate (which is charitable, I know), this means the feeding pool, in the worst case scenario, should always include around 406 individuals. Very similar to OP's estimate of 372.
Let's now move to the population part. I assumed that:
It follows every year 7.4 adults come out of age.
This means (drawing on ideas presented in other responses to this threas) that we need 7.4 new teenagers coming of age each year.
But what about the premature mortality rate of adults in the feeding pool?
My only available data are for the WHO EU region. This region is useful as it includes countries with a great deal of variation (think of comparing Norway with Russia). This should give us the best shot at an estimate.
Now, I only have the following info:
- Age-standardized overall premature mortality in people aged 30-69
years for four major noncommunicable diseases (cardiovascular,
cancer, diabetes, chronic respiratory) = 379.62 per 100,000 (Source: WHO)
Age-standardized mortality rates from all external causes and
injuries = 49.93 per 100,000 (Source: WHO)
Although the former of the two datasets considers 30-69 and not 16-69, I am happy to go along with both of them, for sake of simplicity.
The result is 0.0043% premature deaths every year.
This means that, every year, we have to replace another 2 adults in our feeding pool.
Possible biases: Premature deaths mean less adults reaching the aforementioned ceiling age (70 y.o.); I did not account for this. Moreover, one could argue that the premature deaths factor in diseases that make someone unsuitable to donate; that is, I should have considered only death from external causes as an agent that drains my feeding pool. If that was the case, with a charitable 49.93 out of 100,000, I could expect as small as 0.0005% of the feeding pool disappear by accident (that is, negligible). Our Vampire, however, is not a doctor and does not diagnose patients (I assume); thus, to play it safe he should consider the possibility of feeding from the wrong people or his feeding pool draining faster than he think.
Therfore, I argue that a good margin of error (in excess) is my combined number of premature deaths for all causes. For a population of 406, this number is 2 per 1,000 people. Hence, considering those who turn elderly, overall yearly replenishing rate of the feeding pool should be of around 9.4 individuals per year (again: to play it safe).
This falls within the birthrate range of the countries with the lowest birthrate in the world (see this chart for an overall perspective).
All data, then, point at a community with:
- Around 1,000 people
- Birthrate of at least 9.4 per 1,000 pop.
- 406 potential donors
We can test this by doing a reverse calculation: considering that suitable donors are estimated to be around 37% of the population (all citizens, including kids) (Source data is from the USA, which is definitely not top notch in terms of overall access to healthcare and overall security). Note: some respondents to this thread mentioned figures as small as 3%. This is wrong, because they are looking at the percentage of people who actually donate, which is a free choice. In our case, the Vampire does not act like the Red Cross, so that our 37% will have little choice and every candidate will be exploited.
If 406 donors are 37% of our overall population, it means that, well... our total population has about 1098 people (including the kids and the elderly). This is very close to our expectations in terms of birthrate and replenishment of our feeding pool!
- Total pop.: 1098+
- Feeeding pool (worst case scenario): 406
- Kids in the community (aged 0-15): 151
- Remainder (too weak to donate and/or too old): 541+
- Blood Points needed per year: 933-934 (according to my last simulation)
- Peak individuals out of the feeding pool at any given time (due to short, but not too short recovery window periods): 203 (according
to an "awakening-scenarios" simulation that was run 100 times)
Now, this should set OP up for a situation in which overfeeding should be avoided while, at the same time, account for a sufficiently thrilling long-term survival scenario.
Discussion: communities and their demographic history
Some results might have to be adjusted. After all, I used nationwide data, but we all know birth rates vary regionally. An isolated but thriving mining community might have a birth rate that is higher than the rest of the country.
Also, you should factor some history in the picture: how long has the vampire been there for? Human communities thrive and decline and hardly remain stable. If he came to the mining town in the 60s, he might have witnessed a pop. boom, which would have set him up for the next decades. History is a player even if you consider role play set in contemporary eras... A lot of things have changed on Earth since 50 years ago!
Now, talking short-term illnesses, I did not factor them in. It all depends on the type of community you want. Isolated might mean poor access to healthcare, but it might also mean its inhabitants are shielded from pandemics. Another thing I cannot factor in is whether or not the vampire is a disease carrier. In such case, influenza would spread very quickly among people.
Finally, I would like to invite you to consider the possibility of spiking "death rate from other causes" in order to give your dude a 10 BP boost every now and then. In Russia, which has a murder rate of 10 per 100,000 you could afford a vampire-led murder every 10 years. Also, you could feed on people close to the end of their life cycle to avoid taxing your feeding pool. This should give the vampire some occasional boosts.
I hope this answers your questions both demographically and narratively.
P.S.: I would like to thank everyone who helped me edit this. I came back to update sources and calculations. I apologise if my earlier contribution was not that strong, but I typed it on mobile, with some limitations to formatting.