Geopolymer Concrete, the perfect seasteading material

(Anenome) #1

Geopolymer concrete is a very promising material for seasteading due to its high strength, low-cost, and extreme resistance to seawater, unmatched by any other kind of concrete. It’s a potential broad-basis Portland-concrete replacement material, especially for projects involving molds or slipcasting.

No reason why geopolymer structures can’t last hundreds of years at sea.

I have not yet perfected the geopolymer formula, though I have learned a good bit about what to do and what not to do. I plan to put these into a short monogram and release it for everyone to try.

It was very difficult for us to discover and vett the formula but I’m quite willing to share.

(^A small block of 5,000 PSI geopolymer concrete I produced.)

Let me dig out my notes here…

These are the proportions by weight for our at ~5,000+ PSI geopolymer concrete. These proportions are for a 6,000 grams batch.

  • 101.8 grams of 14-molarity solution Lye (sodium-hydroxide). (This means 41g of lye and 60.7 grams of water). Be careful when mixing this together. Start with a plastic cup of water, 60.7g of it, and then add about half the lye. It will heat the water almost to the boiling point. If you see bubbles forming that’s okay, just stir and let it cool. Once it has cooled a good bit, say 5 minutes or so, add the rest of the lye and stir until it dissolves as well. If you dump in all the lye at once it can boil and sputter and send caustic lye back at you, and it will burn you. If it burns you, wash the spot with water for 10 min. And be careful, because lye can burn your skin in such a way that it will do damage long before you feel any pain, so be careful.

This is the only dangerous step in making geopolymer concrete, and it’s about as dangerous as making soap, which also uses lye.

  • 255.7 grams of Waterglass (sodium-silicate).

  • 15.15 grams of superplasticizer. (Geopolymer concrete turned out to be plastic enough on its own that we omitted this from future batches as unnecessary. It’s generally fairly loose. This is one of its problem! Makes it hard to prepare for spraying and plastering, but perhaps with the addition of nylon fibers it can be made thicker.)

  • 1848 gram of mixed aggregate (sand and 7mm gravel). One point on this, we began ommitting the rock and using pure sand and still obtained a high strength value, but I suggest you play around with the ration of rock to sand and try to find a good medium point. We cut back on aggregate compared to the first pour because the first pour was extremely rocky and wouldn’t even fill the mold we had. The first pour had 1715g of rock and 734.3g of sand. This mix with all sand and no rock came out very beautiful and strong, but it could be made stronger with some rock most likely. This would be a good thing to try out. Also, this rock and sand should be measured out at its wet-weight, not dry weight. So make sure it always has some water in the bag to keep it hydrated. Otherwise dry aggregate will suck water out of the alkali-activator and possibly cause a failed pour when you begin to mix them together. One more note, do not use beach sand, you want some kind of granite-sand or mason-sand. Don’t use beach sand, it results in significant strength loss.

  • 1013g of type-F, low-calcium flyash.

  • 41g of water. One thing we learned was to not play around with the water ratio. You can’t make geopolymer thicker or thinner by adding or taking away water like you can with normal concrete. Instead this will cause the chemistry to fail. The chemical ratios have to be kept fairly consistent. That’s why I say try nylon fibers as a thickener rather than trying to play with water ratios. We did a lot of playing with water ratios and had a lot of failed pours that failed to set-up.

Mixing Process:

  1. Measure out and combine the damp aggregate (sand, rock) into a plastic bucket (do not use metal bucket). Measure 41g of water add it in. Mix the sand and rock for several minutes until everything is well uniformly wet and mixed using a mechanical stirrer of some sort.

  2. Measure 60.7g of water, put into a plastic container.

  3. Measure 41g of solid lye pellets. Don’t leave these standing in the air too long because they will absorb moisture from the air and become gummy.

  4. Pour about half of the lye into the water and mix with a wooden stirrer. Allow the lye to cool down as you mix, then add more lye until it absorbs. Be careful not to add so quickly that it begins to first bubble and then boil. You should be able to feel the heat on the outside of the container and can use that to judge. If mixing large batches of lye solution you will need to mix these the day before and allow them to come down to room temperature before continuing. Cover the lye solution and continue.

  5. Measure out 255.7g of liquid waterglass (36.5% sodium-silicate, 62.5% water). Immediately add it to the cooled lye-solution and stir together.

  6. Pour the solution into the aggregate and mix for several minutes with a mechanical mixing paddle. We used an aluminum-tipped mortar mixing paddle on the end of a drill. The lye will off-gas hydrogen if it comes into contact with just about any metal, but we felt that once it was mixed in with the flyash and aggregate that it wouldn’t be as active against the metal. The alternative was to try to coat the paddle somehow, and that wasn’t a good option as we thought it would surely wear off into the mix. A tough and strong plastic-coated paddle would be idea.

  7. Spray the molds with Pam cooking spray as the mold release (or use any similar mold release, but don’t use petroleum jelly, it’s been known to interfere chemically with geopolymer).

  8. Let it sit for a few minutes, then pour the mix into a mold. I suggest wooden or silicone molds that can survive the heat of curing. We used 2.5" cube molds made of wood and previously coated in silicone caulk. Note: ideally you would de-gas the mix in a vacuum chamber to get rid of any entrained air before pouring.

  9. Cure the geopolymer in a pre-heated oven at no more than 200° Fahrenheit. Any hotter and it will negatively affect the strength. At 200°F it cures in 4 hours. At 85°F it will cure in 24 hours. Any analogous range and length between works too (ie: you could try 120° for 12 hours). It does not need to be covered or kept wet while curing.

  10. Remove from heat when the time is up and remove from the mold (further heat will not hurt or help it). It is now cured and has about 90% of its final strength. Within 3 days it will have 95% of its full strength, and 99% within a month.

A note about flyash:

You can order a flyash type-F sample from Boral free of charge. However if you’re ever in doubt there’s a simply test you can perform. If the flyash is high calcium, it will heat up when mixed with a little bit of water. Calcium compounds in both concrete and type-C high-calcium flyash are what cause both concrete and type-C flyash to cure themselves by generating their own heat, what’s known as the heat of hydration.

If you add a bit of water to a good amount of flyash (say the size of a cup) and it stays completely cool, then you have a low-calcium type-F flyash that is possibly a good fit for this recipe.

If you have a choice, the lower the calcium content the better. 2% calcium flyash is about as good as can be hoped for. I performed this recipe with 5% flyash that was available to me.

Good luck!

And just so there’s no confusion, I am releasing this info under the MIT license:

The MIT License (MIT)

Copyright © <2014> <Michael Eliot, Andy Thomas>

Permission is hereby granted, free of charge, to any person obtaining a copy of this document, to deal in the document without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the document, and to permit persons to whom the document is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.



You can find my proposal for a geopolymer-based Maran floathouse here.

Breakaway Civilization | Seasteading | Ocean Colonization | Advanced Oceanic Cities | Atlantis | Enlightenment | Oceanic Business Alliance | next big thing in business
UK Physicist wanting to learn more about seasteading
Oceanic business alliance | key player network | ocean colonization | big five | get connected | get invested
Open Source Project Leadership
Concrete that thrives in the Ocean
UK Physicist wanting to learn more about seasteading
Concrete Reinforced Geodesic Design
(Jonas Smith) #2

From “Advanced High-Performance Materials for Highway Applications: A Report on the State of Technology, Chapter 3, Candidate Concrete Materials”

Geopolymer Concrete


The term geopolymer represents a broad range of materials characterized by chains or networks of inorganic molecules.1 There are nine different classes of geopolymers, but those of greatest potential application for transportation infrastructure are composed of alumino-silicate materials that may be used to completely replace portland cement in concrete construction. These geopolymers rely on thermally activated natural materials (e.g., kaolinite clay) or industrial by-products (e.g., fly ash, slag) to provide a source of silicon (Si) and aluminum (Al), which are dissolved in an alkali-activating solution and then polymerize in chains or networks to create the hardened binder. Some of these systems have ancient roots, and have been used for decades, often being referred to as alkali-activated cements or inorganic polymer cements. Most geopolymer systems rely on minimally processed natural materials or industrial by-products to provide the binding agents, and thus require relatively little energy and release minimal amounts of CO2 during production. Since portland cement is responsible for upward of 85 percent of the energy and 90 percent of the CO2 attributed to a typical ready-mixed concrete, the energy and CO2 savings through the use of a geopolymer can be significant.
The major drawback of current geopolymer technologies is their lack of versatility and cost-effectiveness compared to portland cement systems. Although numerous geopolymer systems have been proposed (most of which are patented), most suffer from being difficult to work with, requiring great care in production while posing a safety risk due the high alkalinity of the activating solution (most commonly sodium or potassium hydroxide). In addition, the polymerization reaction is very sensitive to temperature and usually requires that the geopolymer concrete be cured at elevated temperatures, effectively limiting its use to precast applications. Considerable research is underway to develop geopolymer systems that address these technical hurdles, creating a low-embodied energy, low-CO2 binder that has properties similar to portland cement. In addition, research is also focusing on the development of user-friendly geopolymers that do not require the use of highly caustic activating solutions.


Currently, geopolymer concrete has very limited transportation infrastructure applications, being primarily restricted to international use in the precast industry. A blended portland-geopolymer cement known as Pyrament® (patented in 1984) has been used for rapid pavement repair, a technology still in use by the U.S. military along with geopolymer pavement coatings designed to resist the heat generated by vertical takeoff and landing aircraft.
Potential applications of geopolymers for bridges include precast structural elements and decks as well as structural retrofit using geopolymer fiber composites. To date, none of these potential applications is beyond the development stage.


Benefits to be derived from the use of geopolymer concrete fit squarely into enhanced sustainability through increased longevity and reduced environmental impacts. The geopolymer systems under development for transportation infrastructure possess excellent mechanical properties and are highly durable, and therefore would result in increased longevity when used in harsh environments such as marine structures or pavements/structures exposed to heavy and frequent deicer applications. Furthermore, these systems rely on the use of industrial by-products (e.g., fly ash, slag). Most significantly, the widespread use of geopolymer concrete would significantly reduce the embodied energy and CO2 associated with the construction of concrete transportation infrastructure, significantly reducing its environmental footprint.


The cost of geopolymer concrete is unknown, as it is still under development. The raw materials are not expensive and the equipment needed for geopolymer concrete is similar to that used to produce and handle conventional PCC. Systems need to be developed that are more user-friendly and less hazardous and that ideally can be used for cast-in-place applications at ambient temperatures. One concern is that many of the geopolymer systems that have been developed are patented, which will increase the cost of implementation.

Current Status

Research into geopolymer applications is at a fever pitch, from small startup companies to major international efforts. Australia and Europe have led significant past research efforts, but there has been a dramatic increase in research in the United States in recent years as interest in developing low-CO2-emitting cementitious binders continues to grow. The first transportation application will likely be from the precast industry, but as of yet, there are no known producers of precast geopolymer concrete in the United States.

For More Information

Davidovits, J. 2002. "30 Years of Successes and Failures in Geopolymer Applications - Market Trends and Potential Breakthroughs." Proceedings of Geopolymer 2002 Conference. Melbourne, Australia.
Davidovits, J. 2008. Geopolymer Chemistry and Applications. Institut Géopolymère. Saint-Quentin, France.
Hardjito, D., S. Wallah, D. M. J. Sumajouw, and B. V. Rangan. 2004. "On the Development of Fly Ash-Based Geopolymer Concrete." ACI Materials Journal. Vol. 101, No. 6. American Concrete Institute, Farmington Hills, MI.
Lloyd, N., and V. Rangan. 2009. "Geopolymer Concrete - Sustainable Cementless Concrete." 10th ACI International Conference on Recent Advances in Concrete Technology and Sustainability Issues. ACI SP-261. American Concrete Institute, Farmington Hills, MI.
Rangan, B. V. 2008. Fly Ash-Based Geopolymer Concrete. Research Report GC 4. Curtin University of Technology, Perth, Australia.
Rangan, B. V. 2008. "Low-Calcium, Fly-Ash-Based Geopolymer Concrete." Concrete Construction Engineering Handbook. Taylor and Francis Group, Boca Raton, FL.
Tempest, B., O. Sanusi, J. Gergely, V. Ogunro, and D. Weggel. 2009. "Compressive Strength and Embodied Energy Optimization of Fly Ash Based Geopolymer Concrete." Proceeding: 2009 World of Coal Ash Conference, Lexington, KY.
Van Dam, T. 2010. "Geopolymer Concrete," FHWA TechBrief, Publication No. FHWA-HIF-10-014, Federal Highway Administration, Washington, DC.

(Jonas Smith) #3

It looks like slipforming with geopolymer concrete could be problematic, due to the requirement for elevated temperatures during curing and having to deal with highly alkaline activating solutions.


To bring this full circle, this TSI blog entry was posted today:

Geopolymer Concrete: The Stuff Seasteads Will Be Made of?

It includes a podcast and a link to the “The Case for Geopolymer Concrete in Seasteading” which is published on the TSI website.

The blog entry also links to this forum topic.

(Anenome) #5

Elevated temperatures can be achieved a number of ways, hot water, hot air, or electrical heating. The actual temperatures required aren’t extreme. As low as 85°F or as high as 195°F.

Geopolymer does require more care in preparation and handling than Portland cement, but that’s not so much a problem for us as for the existing concrete industry which is used to doing things a certain way and resists change out of pure inertia. For what we want to do, its advantages far outweigh the slightly increased cost of the material and care in preparation.

Regular concrete works on a simple 3-2-1 formula: 3 sand, 2 cement, 1 water–you cannot do that with geopolymer, it all has to be weighed out fairly close to ideal for the chemistry to work.

And yeah, a high alkaline solution is involved–again more of a problem for the concrete industry than for our purposes. The major risk is when adding lye to water; add it too fast and too much and it can boil over. But I found it very easy to avoid. The minute that small bubbles begin forming (close to boiling), you let the solution cool down for awhile. This could actually be mitigated by mixing with chilled water or actively chilling the water as you mix, such as in an iced-surrounded bucket.

Once the alkali solution is added to the aggregate and flyash, it’s a relatively harmless mixture. The ph comes down overall a good bit and it can be handled by hand without significant risk of chemical burn. I handled plenty of it bare-handed without any hint of damage.

So overall it’s not a big deal to prepare and mix! And the advantages for seasteading are enormous in comparison to the added difficulty over concrete.


What’s the price of it compared to Portland cement?

(Jonas Smith) #7

Yes, but can you use those methods to achieve those elevated temperatures in a slipforming process? I agree with you 100% that geopolymer concrete seems far superior to Portland, especially in a marine environment and coupled with the basalt fiber rebar. But it’s one thing to use it in precast construction, and another to do slipforming.

I have the same concern as above when it comes to slipforming. When you are creating precast structural elements you have far more control than you do in massive slipforming operations.

(Wilfried Ellmer) #8

Geopolymer Concrete, the perfect seasteading material…

doubth that but ready to see your pilot projects ? any ?

(Anenome) #9

It’s about as expensive as a high end concrete, around $150 / cubic-yard, and basalt is as expensive as stainless-steel rebar. So, more expensive than Portland but not overly so. But considering the performance difference against seawater and ocean conditions, I think it’s far cheaper than concrete for any actual application. And basalt has a number of advantages over steel rebar, for one its considerably stronger, and geopolymer forms a chemical as well as mechanical bond with it, not just a mechanical bond as with concrete and steel rebar.

Just the Maran floathouse project that you’ve already seen.

I do have another project more appropriate for after we’ve got a seastead going, the Turitella floathouse.

(Anenome) #10

[quote=“i_is_j_smith, post:7, topic:240”]
Yes, but can you use those methods to achieve those elevated temperatures in a slipforming process? I agree with you 100% that geopolymer concrete seems far superior to Portland, especially in a marine environment and coupled with the basalt fiber rebar. But it’s one thing to use it in precast construction, and another to do slipforming.[/quote]
I’m better I can do it, yeah. It will require building a skirt around the slipform to hold water. Another option is to fill the interior of the structure with hot water as it slips. That would be a lot of water so some filler material would be used too, something like styrofoam or the like, cheap, bulky, and reusable.

Slipforming and casting in a mold are very similar processes actually, I don’t foresee any major problem with actually making the concrete. Once you’ve got the chemistry proportions down it’s just a matter of weighing out into proportions and mixing it together. This can be done continuously. It’s just you actually do have to weigh it out rather than the more kludge method of water, sand, and cement where the proportions barely matter.

Weighing in large quantities can be turned into a rapid production process, not too worried about that. If you can weight and mix a few hundred grams at a time, little prevents you from weighting and mixing a hundred pounds at a time. In fact it’s less sensitive at those levels as a mistake in grams is easier to make than a mistake in pounds.

(Jonas Smith) #11

I like your thought of using “electrical heating”. If the slipforms themselves were embedded with heating elements you could use the forms themselves to heat the mixture. Not sure how that would work but it might be more viable than filling something like a Troll A leg with hot water! :wink:


Very interesting, thank you for the details. Do you happen to have some estimates of the tensile strength of the material compared to steel and steel reinforced concrete?

Also, would you be available to assist me in building a scale model of one of my designs in this material? Lets say, 5’ x 3’ so I can do few lab test and maybe a water tank-testing?

(Anenome) #14

Sure, do you have an illustration of the design? What production process do you want to use?

The major issue with a large scale model is the ability to get your hands on enough flyash.

A flyash supply chain does exist, but Boral, who supplied me with the low-calcium flyash I used, will sell you a minimum of 27-tons of flyash for $810.

It would probably cost more to ship that to me than to buy it.

So instead I had them send me a 5-gallon sample, which they did, for free.

That’s enough for the small testing I was doing, but probably not enough for your planned 5’x3’ model.

Maybe we can find someone with the ability to store and ship flyash willing to sell it to us at smaller scales for testing purposes. Otherwise, when I fundraise for a scale model I’ll develop that capability myself and at that time will sell flyash at any quantity to people who want to try it out, like you.

(Anenome) #15

Well the problem with electrical heating is that it’s so subject to hotspots where the heating elements are, leading to uneven heating, which is slightly undesirable. If I have a skirt of hot water it will tend to equalize its temperature widely, and it lets me use simple and well-understood plumbing systems, commercial hot water heaters, etc., to precisely control the temperature.

It will probably be enough to fill the slipcasting skirt with hot water on both sides, the heat capacity of water is extremely high compared to the thin sheetmetal the slipcasting skirts will be made out of.

The only reason to fill the inside of the structure with water is to weigh it down for dipcasting when producing the structure on the sea itself, slipcast the structure at sealevel and then allow it to descend into the sea as it is being formed.

(Wilfried Ellmer) #16

in concrete technology curing sistems work with steam - both heat and humidity are required

(Anenome) #17

Not when using geopolymer concrete.

Anything over 195°F yielded decreases in strength, so boiling water is about where I want to be, maximum. The preferred curing temeprature range is fairly low, between 85°F and 195°F.

You cannot take what is conventional wisdom in Portland cement and apply it to geopolymer. They’ve got to be treated as separate materials.

Using steam on geopolymer would be a bad idea, and adding humidity would also be bad, as geopolymer is sensitive to changed moisture ratios, too much water or too little can throw the chemistry off. Geopolymer is vastly more sensitive to proper water ratio than Portland, because geopolymer doesn’t offgas water, it actually chemically incorporates it into its chemical matrix, whereas Portland cures by offgassing water, not incorporating it.

Reviewing DeltaSync’s construction estimates on the simple square Concrete Caissons

I just realized that when scaled down, the thickness of the hull will be very thin,…Also, to start with, we have to look into what hull thickness the structure should have in a real construction situation, keeping in mind “seasteading needs” (meaning to be in the water “indefinitely”)

For example, assuming a 200’ LOA structure and a rebar reinforced geopolymer concrete methods, should the hull thickness be 3", 5", 10" or what? (assuming an offshore capable with a high degree of strength hull)

Also, as a general question, can geopolymer concrete be used in a ferrocement construction method? The reason I am asking is because I envisioned a cheaper method of building for baysteading applications (when strength is not such an important factor). The idea would be to build floats out of marine plywood and than ferrosheating (encapsulating) them in a thin geopolymer concrete layer for the main purpose of reducing maintenance costs.

(Anenome) #19

My only reference point to go off is the concrete sailboats of the 1970’s, which sailed around with about 2" of concrete generally and were designed to flex with the waves just a little bit, not be monolithically static and unmoving like we typically imagine concrete structures to be (another plus for more flexible geopolymer).

This requires internal support ribs, which I haven’t talked much about yet, but they’re much less obstrusive that bulkheads in a wooden/fiberglass sailboat. And since we’re dealing with a tubular structure, I want to see if I can cheat a little bit in how I build them by not slipcasting them in, but adding them in afterwards. However if that proves not to work, I have an alternate plan that makes them slipcastable with a bit more effort.

So the real answer is that we need some professional engineering done on the hull shape before we begin building, and this is not something I’m personally qualified to do. Part of the building process must be to consult an engineering firm and finalize specs.

But because of the concrete sailboat experience back in the 70’s, and the concrete sailboats that sailed literally millions of miles around the planet, including circumnavigations, with a mere 2" of concrete and concrete ribs such as I’m talking about, I feel safe saying we should be able to get away with 3" or 4" of concrete plus concrete ribs, a thickness that the concrete sailboat guys would’ve said was far too thick to obtain certification.

In one instance, a couple made a concrete sailboat 4" thick, and cast bunkbeds into the hull, etc., and the weight went up so high compared to the displacement that the author was unwilling to certify it for ocean-use. He told them to hire a jackhammer crew and reduce the width! You can just imagine the heartbreak of jackhammering to pieces your brand-new boat, but they did. Took 2" off of it via jackhammer, smoothed it back up again, must have have millions of microcracks all over the place needing filling, but it was put back into shape, certified by another professional, and sailed successfully thereafter.

So, 200’ long structure? You might indeed want a thicker hull for that, and a better distribution of ribs. Part of my design is to incorporate the floor mount and keel-weight into the slipform, and use these thicker sections as lateral strengthening-ribs. Then use these two structures to place strengthening ribs after the fact. But we’ll need an engineer’s professional opinion ultimately.

can geopolymer concrete be used in a ferrocement construction method?

If you mean can it be mortar-sprayed, that’s difficult at this point. I would urge you to try pouring the material and see how it moves. I consider it too thin and unsticky to mortar-spray, it simply won’t adhere to a vertical surface. It is a very wet material as cement goes, even though it has very little water in it. And you cannot reduce the water level to make it thicker, as that throws the chemistry off and it will fail to cure adequately.

I have read that one company produced a thick and sticky mortar-sprayable geopolymer formula with the addition of microfibers. I haven’t played with this technique yet but it may be the answer to your question.

Although, if I was going to use microfiber, I wouldn’t use the more common plastic or fiberglass fibers, but rather basalt fiber, which you can obtain from Sudaglass, chopped and ready to mix into concrete.


I was thinking along the same lines of the 2" hull ferrocement hulls of the 1970 and it seems that 4" would be the proper thickness for a 200’ LOA structure.

For such structure (200’ LOA) I am thinking of a continuous pour with collision and structural bulkheads from the same material, or, maybe a prefabricated components construction. My design doesn’t have “curves” in order to simplify the construction process.

Here is a basic blue print shown @ 200’ LOA.

(Anenome) #21

A torus-like design isn’t a bad design. I’ve been thinking about that one as well.

But I think you might think twice about making it blocky. Any blocky design is going to be weaker than a compound or singular curve, and therefore require much more concrete and much more reinforcement to achieve the same strength, and therefore be much more expensive.

You might need 1’ thick walls and thrice the rebar with the shown design to achieve the same strength as a few inches of concrete using a curve-based design.

And while you say this is to simplify construction, it actually could prove to be much more expensive and much less simple than what I’m about to propose to you.

What I suggest is that you look into dome construction techniques using inflateable air-forms. With an inflateable airform you can cheaply build a torus design. Toruses of arbitrary size are possible.

It would come out looking a bit like this:

For a 200’ diameter you’re probably talking a couple floors in a structure like that. The torus design is self-reinforcing and strong, and includes that native moonpool protecting against center-flexing in wave-action.

Toruses should tend to be pretty damn stable in waves, and the larger the better for Draupner-safety.

If you definitely want accessibility to the center, as shown in your design, that’s also doable.

But I think you could achieve substantially the same thing as your design here with two large Maran floathouses side-by side, with a dome-structure built on top of them instead. That would substantially replicate what you’re showing in that design while having a number of advantages, including cheapness to build and no need for mold building. Let me whip up a quick mockup…