Let's say this setup being adequate, 1 civilian hospital severing 10,000 people in the area need to stock 1,000 bags of blood including 1 being Rh-.
In war time, there need to be 10 field hospitals severing 10,000 soldiers which need 1,000 bags including 1 bag of Rh-. It is numerically adequate. But splitting that 1 bag Rh- among 10 hospitals saves nobody. Transferring 1 Rh- soldier from a field hospital without the right blood to another one may be prohibitive or too late. To guarantee Rh-, you will need 1 Rh- in every hospital, that increase the logistic demand 10 fold considering you have to collect 9 times more Rh- than actually necessary and 1/10 available among the population.
I am aware of ethnic minority soldiers from western China who may have a higher probability of being Rh-. But they are mostly put into units that are predominantly from the same or similar ethnic background such as ethnic platoon and company, therefor they may receive special arrangement (special tag perhaps and field hospital in the area with higher Rh- stock). For the rest and majority of Chinese (Han or Mongol etc.), maybe Rh- would not be recruited to avoid logistic problem.
Where did you get 1 unit of Rh- blood per 10,000 soldiers from? Yes, that is clearly inadequate. Even 10 ten times that much is inadequate. On the other hand, making an argument this way is like saying that spreading one MRE amongst 100 soldiers in a company is clearly inadequate. Also, your assumption that civilian hospital blood banks don't transfer blood from other hospitals/blood banks due to this being cost- or time-prohibitive is definitely wrong. In fact for rare blood types this is the USUAL method of obtaining blood.
There Are Rh- PLA soldiers. Here's a news about a military hospital seeking for A- blood donor on social media to save a soldier from heat apoplexy.
Also accroding to multiple comments online, it seams that most hospitals in China, including military hospital, do no store a lot Rh- blood in their blood bank, instead they will keep in touch with the Rh- blood donors for emergency.
So, I did some digging around on the internet and at my local hospital (a 220 bed complex). Here is an Excel spreadsheet which lists the ABO/Rh blood types in the first column, the worldwide prevalence of various blood types in the 2nd column, then China and the US prevalences in the 3rd and 4th columns, and finally the number of units present at the local hospital (on average) for each blood type.
As you can see, both O+ and O- blood units are over-represented in the blood bank compared to general prevalence in the population, and for reasons which will soon become apparent. The tech intimated that for B- and AB- blood (and the even rarer blood types), they do not routinely stock them and have to call around and see where they can get these units, if these blood types are needed. I assume that in the case of the average Chinese hospital/blood bank, A-, B-, and AB- will all NOT be routinely stocked, but both O- and O+ will be overstocked (relative to population) like in the US. Incidentally, donors who are O- (and to a lesser extent O+) tend to be significantly overrepresented in blood bank donations because they know they are the universal donors and tend to donate more generously than other people.
The general transfusion protocol involves typing and crossmatching the exact blood types before transfusing the patient. If the blood type is unknown/has not yet been determined and the patient needs immediate transfusion, O type blood will be given. O- blood is reserved almost exclusively for children under 18 and childbearing-age women, and even then they are only given O- until the exact blood type can be determined (usually within 2 units or so assuming they go in at ~1U/hr), after which they switch to the appropriate type. O+ blood is given to all adult males and post-menopausal females until the proper blood type can be determined. Even if the patient later turns out to be Rh-, the risks/benefit ratio will typically favor transfusing over waiting for a match. The reaction doesn't actually even happen the first time an Rh- person receives Rh+ blood (no anti-Rh antibodies have yet formed), and most don't even end up making antibodies to Rh at all despite the exposure.
