At the end of the 17th and the beginning of the 18th century there was a blooming indigo production in Curaçao. A lot of the plantations had planted indigo plants and had built indigo tanks to produce the indigo dye for export to the Netherlands. The dye was in high demand there.
I have simulated the indigo production process at a much smaller scale. The original process has been described in a old Dutch document "Brieven en Papieren van Curaçao en Onderhoorige Eilanden" (Letters and Papers from Curaçao and its subordinate islands). More specific in the "Memorie Om Indigo Zaat te planten, schoon te maken, te snijden, te slaan, en voorts tot Indigo te bearbeijden in manieren als volgt" (Memorandum how to plant indigo seed, clean, cut and beat the plants to produce indigo). This memorandum is part of book 204, part (folio) 39 - 41. Full source reference: Nationaal Archief Den Haag (A.R.A.) 2e WIC toegang 01.05.01.02 inventaris nummer 204, folio 39 - 41 transcribed by S. H. Vautier. The text is published in the Archiefvriend in June 2006. I will not translate the old Dutch text into English but will just quote the transcribed Dutch text. After each part of the original text I will describe the process in English as I repeated it in 2014 on a much smaller scale.
The indigo plant
Curaçao has two kinds of Indigo plants, the indigofera tinctoria and the indigo suffruticosa. Both plants look quite similar; the main difference is in the seedpods. The indigo suffruticosa has bent pods while the indigo tinctoria has straigth pods. The plant in the picture is the indigo tinctoria (straight pods).
I took some seed pods from a plant during one of my hikes and planted these. The plants reproduce easily and grow reasonably fast. I watered the plants daily; after about 6 months I had more than enough indigo plants to start the process.
Indigo tanks in Curaçao
For the production of indigo dye a system of three brick tanks was used. These were constructed from the rocks that were available in the region. Most indigo tanks are constructed from limestone but there are also indigo tanks made from Knip formation rocks (Savonet). The inside could be covered with IJssel bricks. At least the sides of the opening between the tanks was made from IJssel bricks. The inside of the tanks was covered with a watertigth layer made from lime plaster mixed with ground rooftiles.
The build is in all cases more or less similar. The largest tank is at a higher level than the middle tank and the smallest tank is at the lowest level as you can see in this picture of the indigo tank of Klein St. Michiel. In this way the water could flow from the top tank into the middle tank into the smallest tank when needed. Some of the indigo tanks in Curaçao have one or more additional tanks on the side (San Sebastian, St. Michiel). The purpose of these additional tanks is not known. Maybe these were used as temporary storage tanks to speed up the production process. When there was a larger production than could be handled with one indigo tank a second (Malpais) and sometimes third indigo tank (Savonet) was built closeby.
Indigo tanks were constructed at quite a distance downwinds of the plantation house in the neighborhood of water because of the smell of the indigo production and because the process requires a lot of water.
The process - harvesting the plants and filling the rotting tank
Original description:
Folion 40 verso
Als nú de Indigo beqúaam is, om te snijden, soo snijt men het digt aande grond aff, en toesien, dat de slaeven int snijden, de wortels niet úijtrúcken, Y7 dan in dragte off manden, nade rottback gedragen, en gedaan werden malkanderen geleijt, en dan de back met water gevúlt Even tot boven 't wiet, en met stocken op het wiet inde backe gelegt & twee planken daar dwars over de stocken geleijd onder de balcken voort reijse Vant wiet, als het begint te werken, om dat het niet over sal Loopen
When the indigo plants are ready to be harvested they are cut close to the ground leaving the roots intact. The plants are transported to the largest tank, called the rotting tank and put into this tank. The tank was filled with water till the plants were covered. To prevent the plants from rising from the tank during the rotting process wooden planks were put over the plants.
My rotting tank was a large container that normally is used to mix cement. I harvested branches from two plants in total equivalent of about one complete plant; I have put the container on a table to simulate the height difference in the original tanks and filled it with water from my deepwell. Then I placed the plants in the container (I had to cut them to fit) and put a wooden grid over it.
Original text: Als nú de Indigo begint te Werken, als een kúijp met bier soo laat men het werken, tot dat het begint te sacken dat gemeenlijk een Etmaal dúúrt, de Eerste back werkt langsaam, maar als daar twee a drie maal ingerot is, dan werkt wel Eerder
The rotting starts automatically. It produced bubbles after about a day. According to the original text the whole process should take about one day but this text also mentions that a new tank takes more time. My tank certainly was a new tank. So I let the process continue for two days.
About the smell: the rotting indeed produces an unpleasant smell but not as strong as I had expected from the distance that the tanks are built from the plantation house. I had to stand quite close to the tank to smell it. But maybe that has something to do with the scale of the experiment compared to the actual production.
The process - moving the water into the beating tank
Original text: als het dan Een a twee dúijmen gesackt is, dan tapt men het aff, inde slagback, dat dan als men veel Indigo te snijden heeft, wel aanstonds geslagen werd, maar anders laat men het wel een nagt overstaan, Eer men het slaat
When the rotting is completed the fluid is drained from the rotting tank into the second tank, the beating tank. I used a rectangular container, normally used to transport my dive equipment, as the beating tank. To my surprise the water in this second tank appeared to be green-blue and not colorless as another document described. A clear indication that the rotting had extracted the dye from the plants.
This tank is called the beating tank because the fluid has to be beaten to mix it with air. The oxygen in the air binds with the dye in the solution and produces small granules.
Originele text: Als men nú doende is met de Indigo te slaan, soo heeft men een schoon tinne Coppje bij den hand, alsmeede soet olij, en int slaan, als het veel schúijmt, so doet men telckens een klijn wijnig olie inde back, omde schúijm wegh te neemen, en dan schept men telckens, met een klijn Calbasje úijt de back, en doet daar van wat int Coppie om te sien, off het al grijn begint te setten, Eerst komt het grijn heel fijn, en slaat ment verder tot dat het grijn groffer werd, & dan sakt het grijn op de gront, vant Coppje, en het Water staat boven klaar dan Oordeelt men Ordinaris dat het genoed geslagen is, maar geen saak inde weerelt soo onseeker, van met slaan op de Regte tijd op te hoúden, daarom gist men maar, naar dat het grijn geeft en Zett
While beating the fluid a method is needed to determine if the beating produces the expected result (granules of indigo dye). In the original text a small calabash was used to scoop up some of the fluid; then it was put in a small tin cup to check if granules were forming. I used a glass to do this check. And indeed some small particles were forming. But because I have no experience how large these particles are supposed to be I decided to stop beating after a bit more than an hour. Also in the past this was not an exact process according to the original text that says: nothing is so unsure as to determine when to stop so often stopping was based on a guess.
One remark: while I expected that the rotting process was smelly the beating was even more unpleasant. By beating the fluid the smell gets better distributed into the air and you also get part over the fluid over your cloths so you self start to be smelly.
The process - transfer the granules to the catch tank
Original text: Folio 41
Als men nú op hoúd met slaan, so laat men t staan sacken, als t goede Indigo is, sackt het datelijk, altemets (?) laat men het ook wel een nagt overstaan, & smorgens Vroeg afgetapt, als men Wil afftappen, laat men Eerst het water úijt de bovenste pijp afflopen, en dan de tweede pijp, maar men moet agt geeven, datter geen Indigo meede afloopt, 't water afgelopen sijnde Can men inde groote back sien, off er nog water op is, geen Water op sijnde, de Indigo met de Onderste Pijp inde Vangback afftappen
The next step is to let the granules settle in the beating tank. This should happen quickly if the beating was successful. Just to be on the safe side I decided to let it settle for more than a day. The top layer of the water in the beating tank was transferred into buckets and distributed over the plants in the garden. According to the original text there were pipes at several levels in the beating tank to remove the top layers of the fluid. I have never seen these pipes in the indigo tanks that I have visited. Only at San Sebastiaan there were indications of such pipes.
I transferred the remainder of the fluid into another bucket that acted as my catch tank. To see if there was indeed indigo in this remaining fluid I took another sample in a glass. Indeed there were a lot of particles in this fluid.
The process - letting the indigo dye dry
Original text: affgetapt Weesende, inde sacke gedaan, de sacke opgehangen, dat het Water ter deezen daar úijtloopt, en vervolgens úijt de sacken inde doosen gedaan, en soo te laten droogen
I transferred the remaining fluid after letting it settle for another day through a handkerchief. That handkerchief was immediately colored very dark blue. After letting the fluid drain for a short time the particles became a paste. After letting it stand over for a night I got a small piece of indigo dye.
Conclusions
I have been able to reproduce the old process on a small scale. I can confirm that the process produces an unpleasant smell but not only from the rotting process but also from the beating stage of the process.
The process is not very efficient. It took a whole plant to get a small piece of indigo dye, about 1.5 cm in diameter. The whole process takes about 3 to 4 days from indigo plant to indigo paste.
The small scale process has far more losses than the actual production processes will have had. Especially in the last stage where the particles are transferred into burlap sacks (in my case a handkerchief) a lot of the dye was lost during the transfer and in the handkerchief. On a larger scale this loss will be relatively less.
Putting a second and third load into the rotting tank before transferring the fluid into the beating tank also increases the efficiency. More dye will be dissolved in the fluid. Beating longer than I have done could also lead to more granules in that stage. After I completed the simulation I have read somewhere that the beating took several hours. I did it for a bit more than one hour.
But at least it worked; I now have a locally produced piece of indigo dye (probably the only piece on the island) and I have actual experience with this former Curaçao production process.
My indigo plantation at home
Tiny flower of the indigo plant
Seed pods of the indigo plant
My Rotting tank setup
Filling the tank with water from the deepwell
Harvested indigo plant - in total equivalent to about 1 whole plant
Rotting tank filled with the indigo plants - grid prevents the plants from rising
24 hours from the start - rotting has started
24 hours from the start - visible bubbles
48 hours after the start - a greenish blue film on the surface - rotting completed
Time to put the second tank in place - the beating tank, the container of my dive equipment
Fluid from the rotting tank transferred to the beating tank 
Look at the color - clearly the dye is extracted from the plant
After leaving it stand for a day I started beating the fluid
After a bit more of an hour I decided the beating was enough (I was tired and quite smelly)
This sample shows that the beating produces some granules
78 hours after the start - the bottom layer of the fluid is transferred into the third tank, the catch tank
Another sample from the catch tank shows a lot of granules
96 hours from the start - I transferred the remaining fluid through a handkerchief
After draining the water a indigo dye paste is left
Final result after drying - a small piece of indigo dye
My indigo tank consisting of three containers and a bucket with a handkerchief
