Use of Castor Oil and Brake Fluid as a Substitute for LHS-2 in Older Citroëns

Updated January, 2008

Part 1 Part 2 Home

In my last page, I outlined my reasons for finding a reasonable substitute for LHS/LHS-2 hydraulic fluid as used in certain Citroën models. While the use of pure brake fluid has been in widespread in these automobiles for a long time, mostly out of necessity, ride, shifting, steering and longevity has suffered. Many owners of "red fluid" cars have never experienced correct performance.

After learning that the first fluids utilized in these cars contained castor oil, I have tried to determine whether this lubricant could again be employed as an additive. After some reading and probing, I found that there is already a small group of owners who use castor oil, but the amount needed seemed to be, at best, conjecture.

I determined to find the matching ratio of castor oil to brake fluid to match the viscosity of pure LHS-2. After running a series of tests, I found it to be 9.25%. Using that percentage, therefore, one would add 3 oz of castor oil to 1 U.S. quart. Using metrics, one would add 92.5 ml to 1 liter of brake fluid.

To arrive at this figure, I sent a set of 12 samples, each one of an increasing ratio of castor oil to brake fluid, to an independent professional tester, Geoff Byrnes, now retired, of The Coatings Laboratory in Houston, TX. I came upon him more or less randomly, but he quickly became interested in the project, as he had at one time owned a Citroen. He even did the work for free!

The samples were of 400ml. each. 200ml was the minimum sample size, but I chose the larger quantity because it allowed more precise mixing. On July, 2000 I received this data from Geoff:


Amount of castor oil added to 400 ml of brake fluid (ml) Viscosity in millipascal-seconds (mPa. sec.)
Pure DOT 4 Brake Fluid
Pure LHS
Pure Castor Oil
2.5% 10 25.72
3.75% 15 26.4
5% 20 27.4
6.25% 25 29.8
7.5% 30 31
8.75% 35 33.1
10% 40 37.2 (Anomalous, should be ~35)
11.25% 45 37.5
12.5% 50 41.6

If you look over the above chart, you will find that the viscosity of LHS-2 matches that of a sample somewhere between 8.75% and 10% castor oil/brake fluid. This data was further examined by a designer/engineer of rheometers, Jint Nijman. He agreed in principle to the findings but offered a correction to account for some deviations caused by the testing equipment.

It has been determined that use of DOT 3 fluid is slightly less hygroscopic and, though its boiling point is higher when fresh, as it ages the boiling point of DOT 4 fluid drops more quickly. With this in mind, we prefer to use DOT 3.

The discovery of the correct ratio of castor oil to DOT 3 brake fluid is a significant find but ideally, more experiments chould be done. It would be desirable to find the viscosities of the mixture at higher and lower temperatures to find the viscosity index.

If the specific gravity were known, it would be easier to maintain systems at optimum mix ratio, as is done when determining the freezing point ot anti-freeze in cooling systems. For now, it will be necessary to premix. I would advise keeping only as much of the premix on hand as readily needed because once brake fluid is exposed to the atmosphere, it begins to absorb moisture. However, it is possible that the mix may be less hygroscopic.

The recent discovery of original Citroen technical papers have revealed that cars issued with castor oil mix experienced clotting of fluid in cold weather. For this reason, it is very important that those who decide to use castor oil realize this. If you live in a cold climate, make sure you drain your reservoir in the fall, refilling it with pure brake fluid. Then run the car for awhile before driving or laying it up for storage. In the spring, you can add 3 oz. of castor oil. This will be sufficient to bring the viscosity up to a fair average, despite variations in temperatures encountered during the usual operating season (assuming you store your car in the winter). If you do this, you needn't do a routine drain and refill. If you live in a climate where this is not necessary, the entire system should be drained and fresh refilled every 2 years, annually in damp or humid conditions.

Late breaking news: Many thanks to Claude Moritz of France for his work on this. Recent research has revealed that the lubricant used in many automotive air conditioning systems nowadays, PAG (polyalkalyd glycol)  has very close affinity to brake fluid. In its thickest available viscosity, 150, it has shown promise as an additive in early tests. It seems to be far more soluble and seems to solve the low temperature risk, but requires higher ratios to attain viscosity. More work with finding cross brand compatibility between brands of DOT brake fluid and PAG is needed, plus we are in need to some long term in-the car experiences to truly stand behind the use of PAG. All external findings point to no issues with rubber incompatibility.

If anyone has a car they would like to do tests in, the Citroen world would be grateful. If you are interested, please contact me.

Please be aware that using this mixture in your car will be done strictly at your own risk.

Many thanks to Jan Forrester, Geoff Byrnes, Jint Nijman, Stan George, Tony Jackson, Vincent Jammes, Robert Dircks, and many others, whose advice and encouragement has meant so much!

Copyright 2013 by Mark L. Bardenwerper, Sr. All rights reserved.
Last Modified 01/19/13