Karhedron

I would start with a GWR O38/O42 as these are a pre-war design . Despite being different diagram numbers, these diagrams were identical as far as I have been able to tell. The only difference seems to have been that the O38s were built for United Dairies and the O42s for Express (plus the smaller dairies that had opted for the same fittings. The O39s were the same apart from having a different ladder with a central platform. So from a single model with 1 alternate ladder tooling you could produce O38, O39 and O42. This would cover about 50 vehicles for Express, United Dairies/Unigate, CWS, IMS, West Park Dairy, and Nestle. It would cover both the colourful 1930s liveries and the later BR dull silver and (I think) St Ivel. These worked all over the (G)WR and SR. While only representing about 10% of the fleet, it covers all the main iconic dairies and several colourful liveries. If we could stretch to a second prototype, I would opt for the GWR O57 or O60. These were built in large numbers after WW2 and featured the distinctive ladder and platform at one end. The main difference is that the O60s were a BR design and featured roller bearings. So with 2 toolings and one ladder variant, you could build 3 visually distinctive styles of tank covering everything from an array of colourful 1930s liveries to the St Ivel and dull silver liveries of the diesel era.

I have to agree with @rapidoandy. Milk tanks are a bit of a speciality of mine and there are actually a lot more variations than are apparent at first glance. Milk tank chassis did change over the years but the tank fittings are where it really gets interesting. There are central manholes vs end manholes. Some tanks had round ladders, others had square ladders with a platform. Tanks might have 4, 6 or 8 supports between them and the chassis. There might be 2 or 4 straps holding the tank in place. Some tanks had walkways mounted on top. Some chassis had traditional bearings whereas newer ones had roller bearings. To top it off, most of the these features could be found in various combinations with others. Even if you omit detail that is too small to see, there is still more variety in milk tanks than many people realise. So you either produce an accurate model knowing that it is only accurate for a fraction of the tanks in service or some sort of hybrid. If you take the hybrid approach, will the resultant model be any more appealing to detailed modellers than the current offerings from Hornby and Dapol? I have done quite a lot of research along with @BMacdermott and I know what I would choose to produce. But that would just be a level of compromise that would be acceptable to me. Whether it would be economically viable to produce and sell in the volumes required is another matter.

The Long Drag can be viewed here. Appleby is at about 43:00. https://www.yfanefa.com/record/11077

It is currently being laid out so hopefully not too long now.

Depends on what substitue you are using. Almond milk is pretty bad as it is very water-intensive and mostly grown in the US. So there is also the environmental impact of shipping the stuff across the atlantic. I haven't been able to find numbers but I would be suprised if imported almond milk has a lower carbon footprint than british milk. On the other hand, oats grow happily in cold damp climates so oat milk is better on both the water front and carbon-footprint front.

I was following the OP's request for a GWR-ish light dairy train. I do agree that SR stock would seem more likely There is a nice shot here of milk churns being unloaded at Clapham Junction in 1926. Not quite sure what vans they are in. I know the SR often favoured Van Bs and Cs for milk churn traffic.

Technically milk vehicles were not classed as freight but as NPCCS (Non Passenger Carrying Coaching Stock). This means they would have operated more like passenger trains and indeed were often tacked on to the end of passenger trains, particularly on branch lines.

West Park Dairy operated just a single creamery at Market Lavington in Wiltshire. They were taken over by Express Dairies in 1937 so would no longer be operating in 1947/48, although the tanks were still in existence and may have still been in their original livery. Assuming there was a dispatching creamery somewhere on the Bude branch, I can see the tanks being tripped to Oakhampton and then worked through to Exeter. The following dairies were in operation around nationalisation and making use of rail tanks: United Dairies Cow and Gate CWS Express Dairies Ambrosia MMB There was still some churn traffic at this date so it would not be out of the question to see GWR Siphons full of churns (or even a mix of both types of vehicles).

Yes, rejected at the bottling plant. Usually the reason was that the milk has started to spoil. Whether this was due to inadequate cleaning from the previous load or slow transit from the dispatching dairy I am not sure. The tanks were pretty old by this point (I am talking 1970s here) which probably did not help.

Yes, vitreous enamel like an old steel bathtub. There were instructions that milk tanks were not to be loose shunted as rough handling could crack the enamel. Some tanks were lined with Staybright Steel (what we call stainless steel today). I do not know what proportion of tanks had which lining as the tanks stopped carrying descriptions like that after nationalisation.

The manhole was for inspection and washing out. There was a separate small inlet which was used for filling as shown below (although since it is a demo, they are actually using water :D ). The basic design of the tanks is almost 100 years old and it has to be remembered that they were introduced to replace churns. It was a big step forwards in terms of hygiene compared to what had gone before but definitely not what we would consider ideal by modern standards. I have seen several photos from the 60s and 70s (including the one in my previous post) which suggest that filling via manhole may have occurred as an expediency. I have also spoken to the former manager of the Express Dairy in South Morden who recalls that up to 1-in-3 tanks were rejected on arrival. Possibly another reason that the traffic shifted to road.

I have been chatting with Keith Sweetland who used to work at Express Dairy Seaton Junction. He has confirmed that tanks were always filled from the top.

It is possible but Lostwithiel was not involved in local distribution until 1960 when a bottling line was installed. When first opened, Lostwithiel's job was simply to collect milk from the surrounding areas and cool it for dispatch to London. Even more puzzling is that one of the MICAs appears to be parked by the boiler room, you can even see the mound of coal behind it suggesting this is where coal wagons were normally unloaded. That seems to be to be a very odd place to load or unload food products.

I am afraid I don't know what the MICA was doing at Lostwithiel, I haven't really seen much about MICAs used for dairy traffic at all. The GWR favoured Siphons for churn traffic and these were phased out by the introduction of milk tanks (MILTAs). Possibly the MICAs could have been used for butter or cheese but I have not heard of this in practice I think Lostwithiel just dispatched fresh milk to London at this date rather than processing it.