Owens Valley Mining (Part 2) - Bartlett Plant - Columbia-Southern Chemical Corp - Soda Ash & Boron

Soda Ash

Loose Soda Ash - photo from About Mechanics article on soda ash

Most of us live our lives without thinking about how many times we run across a product or interact with something that has been effected by soda ash in our daily lives.  A very good article on how Soda Ash is made can be found at chemicalaldepartment.wordpress.com - the article in part covers "The main consuming industries of soda ash are glass, detergent, laundry soap, sodium silicate, cotton yarn, dyes and dyestuff, paper board and other chemical industries."

"There are two main ways of refining sodium carbonate. Trona, a mineral that occurs in natural deposits in evaporated lake beds throughout the world, is the most common source. Some salt water also contains sodium carbonate compounds. To extract the ash, scientists usually subject the trona ore or salt water to different crushing, heating, spinning, and other techniques." quote from About Mechanics article.  Obviously the production at Bartlett is following the processing of using the lake-bed Trona as the starting point.

To give some idea of the uses of Soda Ash, here's some ratios in 1981 for the use of Soda Ash in the US is as follows from the chemicalaldepartment article:

50% Glass

18% Manufacture of Chemicals

5% Soap & Detergents

3% Pulp & Paper

3% Water & Sewage Treatment

21% Misc (including exports)

In 1981, 7544 metric kilotons were produced per year!  I'm not sure what the ratios and production in kilotons was back in the 1950s, but it gives some idea the diverse uses of Soda Ash

The chemicalaldepartment article mentions that in the US, most of the Soda Ash produced is from natural sources, including refining from naturally occurring Trona minerals.  One of the production points where Trona is harvested from the Owens Lake brine and shipped from Bartlett.  Additionally, Soda Ash, Pot Ash, and Trona mineral is also produced at the plants at Trona, Borosolvay, and West End along the dead shoreline of Searles Dry Lake, just to the south east in the next valley over from Owens Valley, where the Trona Railway ships it out by rail connection at Searles Station on the southern part of the SP's Jawbone branch.

Columbia-Southern Chemical Corporation - Bartlett, Calif.

After diving deeper into the history of the long and complex story of soda ash processing on the Owens Lake, here's a short quote for Columbia-Southern, from the wiki.

Columbia-Southern Chemical Corp wiki
"The plant perfected their sodium carbonate recovery process from the Owens Lake brine, and built the full-scale Pacific Alkali Co. soda ash plant here in 1928. In 1944, Columbia-Southern Chemical, a subsidiary of Pittsburgh Plate Glass Company, purchased the plant. The plant continued to use the old sodium carbonate recovery methods here until the plant was upgraded late in 1958 to incorporate the sesquicarbonate process, and it had three times the capacity of the old plant."

Bartlett town site, about a mile north of the Columbia-Southern plant - Tales Along El Camino Sierra group

Also check out this great website of Owens Valley History!  There's a lot of good information which I'm still trying to digest there.  One part of the website shows a map with Columbia-Southern Chemical at Bartlett, with US Hwy 395, CA Hwy 190, and the SPNG removed, suggesting this map was post-1960 vintage.  Some of the photos of Bartlett (townsite) are pretty good too, which I didn't know about until I saw the pictures... Still so much more to research!

In my previous post Jawbone Branch Part 3 - Consists & More Bartlett Research, I pulled some data from the California Division of Mines for the Columbia-Southern Chemical Corp plant at Barlett that give the following information.  In 1920 the lake level had dropped due to the aqueduct.  Some times the Owens Lake concentration of brine reached 35% within the salt crystalline solid structures - which also changed seasonally, but most of the sodium bicarbonate and sodium sulfate to supply the production of minerals crystallized out.  Some production was then shifted to "dry" salt-cake gathering of soda ash and Trona from the dried lake bed.  Into 1923 and 1924 The Inyo Chemical Company drilled brine wells, sunk into the densest part of the lake bed to concentrate up to process.   Varying concentrations and stubborn problems with filtering the bicarbonate in the carbonating towers and boron also tended to precipitate out at the same time.  By 1932 the Inyo Chem. Co. was closed down.

In 1926 the Kuhnert Syndicate built a test plant at Bartlett using brine containing 9 per cent or more sodium carbonate, which with the controlled application of flue gas with 15% carbon dioxide would force the precipitation of sodium sesquicarbonate (trona mineral) out of the brine.  This new more controlled process didn't run into the same filtering problems as ICC did two years before and the need for lime kilns used in previous processes was not needed.

Bartlett Plant & town overview - Tales Along El Camino Sierra group

The new plant was built with 16 towers, 80ft tall by 5ft wide for the carbonation process.  The well brine was solar concentrated to create a feed brine of about 13% then heated with steam.  The remaining stock was separated in centrifuges from the precipitated trona mineral.  Some of the trona cake was then calcined in furnace, resulting in white soda ash.  Boron was also able to be drawn out of the liquid that was drawn off from the trona precipitate step.  Of interest to the operational modeling of the Bartlett plant is the detail that the solar concentration process, starting around May when the weather warmed up and usually stopped around the first of October.  If this is correct, then this only leaves about 7 months for the production of the 26,000 tons of products.

Changes of Ownership in 1944 

Bartlett 1944 - Fairchild Aerial Survey (RR East to Left)

In 1944 the plant Pacific Alkali Company at Bartlett was sold to the Columbia-Southern Chemical Corporation, a subsidiary of Pittsburgh Plate Glass Company.  They left the plant alone until the late 1950s by which time they decided that the nearly inexhaustible supply of trona at Owens Lake was worth upgrading their plant.  The upgrades came on-line in 1958, tripling the production rates!

Map of Bartlett, Columbia-Southern Chemical plant and Cottonwood Creek. - Calif. Div. of Mines - David Coscia collection (cropped)

As a side note: There are still remains of charcoal kilns several miles south of Bartlett at Cottonwood Creek, so I'm not sure where they're getting their lime from for this process. The charcoal was probably being used for the CO2 needs of the processing of the soda ash and trona at the Bartlett facility and others in the area.

Crunching the Numbers for the Model

I've calculated the shipments based on the California Department of Mines data from 1928 production rates, which supposedly were pretty steady for the 30 years the plant was in operations.  I believe the plant operated 7 days a week (however this may have been semi-seasonally), given the furnace and CO2 production systems needed for the Soda Ash production and purification.  So if loading only on 5 days/week means that my yearly production numbers were divided by 251 days, resulting in the 12,000 tons of Soda Ash equating to 4.6 (50-ton) carloads per 5-day week.  During the same time 12,000 tons of Trona mineral was also being produced, which would lead to 4.6 (50-ton) carloads per 5-day week as well.  I think in the longer run these two 4.6 carloads/week would stack in an alternating fashion, such that one week 5 carloads of Soda Ash would ship and 4 carloads of Trona, then the next week it would reverse with 4 loads of Soda Ash and 5 loads of Trona.

The 40-tons of Boron produced for a 5-day working week (or 7 real days of production) would suggest that the weekly carload would be short loaded or they would have to delay the next carloading for an extra day or day-and-a-half to get a full 50-ton carload.  I could probably increase the apparent randomness of the loads coming out of Bartlett by shifting the Boron shipment to be when the load would actually reach the capacity for a 50-ton car, not a 40-ton car or short-loading a 50-ton car.  If the car was instead loaded and shipped every 8th day of production.

This seems to be a question of scale, and while I could see maybe up to 3x50 tons 3x week of borax... that would put the car loading out of Bartlett at more like 7 cars per trip, which seems a bit high from what the photographs of the Owenyo Local look like during the 1948-1956 era.

Bartlett plant with notes for buildings - Green line is roughly planned edge of layout. - Aug, 1947 Aerial photo set

I've re-run my calculations to convert this to realistic car loadings as follows: These yearly production rates would average of about 33 tons per day, 4.6 (50-ton) carloads per week of Trona mineral, and another 4.6 (50-ton) carloads per week of soda ash, and about 1 (40-ton) car per week of borax.

This means about 10 car loads of mineral loads out per week, or about 3-4 cars per tri-week trip just for Bartlett, which seems possible. Maybe one car in of fuel-oil to dry out the minerals.

Further Digesting the Data

My basic yearly average calculations for about 10 cars per week, of the 12,000 tons of soda ash, 12,000 tons of trona minerals, and the 2,000 tons of boron, but if it actually had to be produced in only about 7 months!  Seven months should work out to about 140 days of loading product.  This means that the 12,000 tons was being loaded at about 600 tons per week or 12 (50 ton) cars of soda ash, 12 (50 ton) cars of trona, and 2 (50 ton) cars of boron per week.  This would total about 26 carloads out per week, Spreading this between the multiple trips of the Owenyo Local per week, this should work out to about 4-5 cars per day, or 8-9 when the local was tri-weekly.  I'm not sure this would accurately represent the production rates.

Newest updated drawing with resized inside runaround and Glass Co loading spots.

However, the possibility that the track arrangement at Bartlett could have had 3 or 4 cars spotted, and ~3 car off-spots stored in the run around track. They may have had the local respot the cars on the northbound trip, or they could have had a cable or car-mover used to reposition the loaded cars into the runaround and pull the next batch of cars up.  So that would seem to suggest a maximum daily loading (between switching moves by the SP) of about 6-7 cars.

Fuel Oil?

There's also a comment about steam heating being used to drive off the water from the brine stock, so I wonder if they could continue production during the cooler months.  So it's probable that they were using fuel-oil to fire the furnace and produce the steam.  So maybe I get to split the difference a bit on the weekly production rates.  Currently I'm thinking, maybe 2/3-3/4 of the yearly production was done during the summer months, and the rest was taken care of during the winter with extra fuel oil heating and some down time for the plant for repairs and maintenance work.

Additional Annoying Evidence

I want to look at a couple of the aerial photos I found.  Thankfully these are dated and so we can do a bit of photo analysis.  I'll probably do a future deeper dive on these photos, combining the same day snap-shots with Owenyo & Lone Pine.

Aerial photo of Bartlett on Friday, 8-1-1947.

Seems there's two tank cars spotted at Bartlett's plant on this early August 1, 1947, a Friday.  The photo appears to be a late morning based on the shadows.  I'll have to cross check if 1947 was still 6-trips/week or if they'd dropped to 3-trips/week by that year.  If this photo is between the nightly trips across the branch, maybe they were loading at night, leaving the plant's spurs empty during the heat of the day?

Slightly more clear 9-10-1955 Bartlett - Aerial photo 1VJZ000060111 (North photo)

Here's a pair of photos taken on 9-10-1955 of Bartlett's plant.  It looks like there's one boxcar.  There are two new towers built over the southeast corner of the main building.  Maybe for covered hopper loading?

A bit blurry 9-10-1955 Bartlett - Aerial photo 1VJZ000060112 (South photo)

The 1955 date would put this past the end of 2-8-2s working on the Jawbone, and RSD-5 or SD9 would have been pulling the local.

In Closing

I'll probably aim for 3-4 cars/day for the 'summer time' rush, maybe up to 5 cars/day.  I don't really want to push to more than that, unless I find that the loading rates of the SPNG dropped off during the summer months.  If the full yearly production of Bartlett was produced only in the summer 7 months, then that would mean about 1/3 of all the traffic on the Jawbone branch would have to be going to Bartlett.  Maybe in August it did get that high, but this still seems a little high.

Once again, one of my favorite photos - SP 3237 switching at Bartlett on the Jawbone branch - Leo Barusch photo - Dani Collection

The problem is right now I don't know, but I'll continue to do research and in the end have fun letting the traffic fluctuate somewhat.  I'll probably continue to play with the shipment schedule spreadsheet and estimated production rates for Bartlett. 

One nice thing to do on a model railroad is keep the car orders from the on-line shippers from becoming too stale.  The "problems" that I'm confronted with now in my research that seem to show inconclusive results could also be taken as a silver-lined cloud, giving me license to have some flexibility in my car assignments and scheduling.

Jason Hill

Related Articles:

Owens Valley Mining (Part 1) - Owenyo Transfer & Commodities - Overview of what was transferred between the SPNG and SPSG at Owenyo.

Jawbone Branch Part 3 - Consists & More Bartlett Research - Very early concept for Bartlett from 2020, Nov 10.

SP Jawbone Branch (Part 42) - Redesigns for Little Lake and Bartlett - 2024, Mar 7 update on layout track plan to be more accurate.

Jawbone Branch (Part 44) - Soapstone - An Unexpectedly Useful Material - Really the first in depth post about mined minerals from the Owens Valley, so I could say this was really the first part of this series.

Jawbone Branch Layout Build Index - Construction of the layout postings.

Owens Valley Mining (Part 1) - Owenyo Transfer & Commodities

I covered the basics of Soap Stone in Jawbone Branch (Part 44), but I think the topic of Owens Valley Mining is specialized enough that the detailed dives into it should split a bit from my regular Jawbone Branch series of blog posts.

SP 3203 & SPNG 9 both working around Owenyo in Feb, 1951 - Martin Hansen collection

As I want to get more into the research of the Owens Valley Mining and what was shipped by rail.  To do this I'll be borrowing the table from Joe Dale Morris' excellent book SP Slim Princess in the Sunset 1940-1960, plus I'll be adding in the research I'm doing for the Jawbone's standard gauge line down the west side of Owens Lake and south through the greater valley towards Inyokern.

SPNG Shipments

SP 102856 boxcar at Spot 11, according to my spot numbering system, on SPNG Owenyo North Transfer Platform (NTP) in 1954 - owensvalleyhistory,com s-l1600_ebay01_sml

Many of the transfers at Owenyo won't be too obvious what exactly was being moved between the NG cars and loading the SG cars.  I hope to have some fun with staging figures with all the sacks, bulk, bagged, piled timber, wheelbarrowed, hand-trucked, or conveyored lading being moved.  The NG also received outside freight forwarder shipments coming in from the rest of the world to the folks on the SPNG villages and mining camps.

List of industries & commodities shipped on the SPNG - table from Joe Dale Morris's SP Slim Princess in the Sunset 1940-1960 book.

Some of the shipments obviously dropped off by the 1940s that I'm modeling, such as the silver ore and bullion shipments, while others, such as the Perlite, didn't come until a couple years after my modeling era cutoff.  Sadly most of the colorful operation of bulk oil movements by rail to Owenyo proper and transfer to the SPNG was abandoned after WWII, as the traffic using the newly upgraded US Hwy 395/6 allowed semi-trucks to haul the fuel in from Mojave and Route 66.

I'm still looking for more information about exactly how much production was shipped by each of the various industries.

Modeling Owenyo Operations

Early development car-spot numbering for my HO-scale Jawbone Branch.

By 1952 the SP was using a station number system.  Owenyo was assigned 3144, Bartlett was 3130 and Little Lake was 3089.  Mojave was 1579.  I'm not sure exactly what the numbering was based on.  It doesn't appear to be the same as the Santa Fe's system which was by straight Mile Post number.  The SP system seems semi mileage based, but with different number series, changing by division.

I will be using the station number system with additional digit at Owenyo to show spot number based on the chart I made above, which is my own fabrication.  Two small changes to the drawing need to be made.  Because the Dump Trestle is going to be shortened to only 4 cars, Spot 5 is now assigned to the Stock Transfer Platform.  The second is that the lower through track in the chart is actually the mainline, so spots 26-30 are not correct to be shown as a place to leave cars, so those spot numbers should be considered void.

Loading chart 2024-04-26, roughed in for total car capacity calculations.

The above loading chart for traffic on the Jawbone Branch is currently just a rough form for me to work out the average train size and how much traffic that will represent on the layout as a baseline.  It seems I could add a bit of traffic or adjust the car counts at Little Lake to have a bit more traffic on Day 1 and Day 5.  Probably in the form of a few extra mineral cars to Owenyo or non-peak season stock car movements.  In reality, these light spots were probably filled by traffic for Lone Pine, but as I won't be modeling Lone Pine (at least currently), I won't have to deal with that traffic. 

I'll be planning to mix up the assignment of cars at Owenyo.  The Transfer Trestle (OTT) probably was spotted and allowed the smaller NG cars to be shoved up and dumped into the standard gauge cars over a couple of days to reach full loads.  There may have also been some separation of either which exact car was shipping bulk soapstone, vs dolomite, vs talc.  It could have also been some days they had one commodity and the next trip it was something else. 

Edit 2024-04-29: There's some new info coming together for Bartlett that indicates the production was at least partly seasonal.  I'm not sure what the production fluctuation, if any, would equate to in terms of carloads being shipped and fuel being supplied by tank cars.  I'll probably adjust the number of cars in and out of the Bartlett plant, but I'll talk about this more in a future post.

More Research to Do

Owenyo Trestle - SG side - unknown calander photo - Ronald Grau collection

Perhaps that's what I'm seeing in the above photo and some aerial photos where an extra set of GS gondolas were off-spotted on the trestle lead track.  The Joe Dale Morris book includes some description of the NG's operations regarding which days of the week where the NG went and collected cars from.  This could give me more info as to which days of the week the transloading was done to the SG cars.  Still more research to do!

In Closing

Hopefully the research on the mining will make my operations on the model more realistic than just shuffling plastic boxes around.

I'm going to wrap up this first of the Owens Valley Mining series at this point.  I'm already starting Part 2 and Part 3 so I don't make each sub-topic too long of a post.

Jason Hill

Related Posts:

Jawbone Branch (Part 44) - Soapstone - An Unexpectedly Useful Material - Really the first in depth post about mined minerals from the Owens Valley, so I could say this was really part 1 of this series.

Jawbone Branch Layout Build Index - Construction of the layout postings.