A proa is a double hulled vessel with all or some of the following attributes:

  • The rig is mounted in or on a hull.
  • The hulls are frequently different lengths and shapes.
  • Steering and leeway prevention are also in one hull, generally but not always, the same one as the rig. Steering can be achieved by rudders, paddles, oars, crew movement and fore and aft movement of either or both of the centre of effort, or centre of lateral resistance. Leeway resistance can be provided by hull shape, leeboards, daggerboards or oversize rudders.
  • They are able to sail in either direction. This generally involves shunting (the equivalent of tacking or gybing, see diagram), although as with most things proa, there are numerous variations of this as well.

All these variations indicate that the ultimate proa is yet to be built.

Proas were probably the first improvement made to a floating log by early sailors. A second, smaller log was lashed to a cross beam to stop the main log capsizing. Rigs were added later when technology allowed. These boats have been scattered through the Pacific and Indian Oceans, where they have been undergoing development for hundreds of years. Consequently they are very highly developed for their conditions of use, and the materials they are built from. They sail in generally warm, trade wind conditions with crew who are not only fit and agile, but are superb seamen. They range in size from small models ballasted by baby coconuts up to large ocean crossing vessels capable of carrying a dozen people and their supplies for many weeks at sea.

However, they are by no means the end of the story. Our ancient Polynesian cousins had no access to epoxy, fibreglass and carbon plus they seldom had to sail upwind in less than idyllic conditions crewed by the wife and kids and driven by the need to be back at work by Monday morning. If they had, then their boats would be far different to what they are.

Western sailors have been slow to adopt proas. This is partly because until now they were difficult to manoeuvre, and partly, I suspect, because many people are unhappy with boats which are not symmetric fore an aft. They received a brief burst of publicity after the 1968 Singlehanded TransAtlantic Race (OSTAR) when the 12m/40′ proa Cheers achieved 3rd place. Cheers was designed by the legendary Dick Newick and was a masterpiece of elegant simplicity. It was sailed by Tom Follett, arguably the best, if not the highest achieving, yachtsman of his time. Cheers was different to traditional proas. It had the rig, rudders and accommodation in the windward hull and looked like a double ended trimaran with the windward float missing. The benefits of this set up are reduced weight, windage and stress of a second set of beams and a third hull and the weight of the rig and foils are all in the best place to enhance righting moment. To differentiate between traditional and Cheers type proas, Newick coined the term Atlantic Proa for proas with rigs, rudders and accommodation in the windward hull. By default, traditional proas scored the prefix, Pacific.

Cheers was schooner rigged, with heavy, wooden, unstayed masts. She was narrow by today’s standards at 4m/13′, had a conservative sail area and was a pig to shunt (more later on this term). She also had inadequate freeboard at the bow, making for dangerous running in big seas or strong wind. Despite these faults, Cheers showed what could be done and started a rush of proas, mainly by the French singlehanded crowd. These 2nd generation boats suffered a rash of misfortunes, mostly caused by sailors running before they could walk. After numerous capsizes and DNF’s, they were banned from short handed yacht races in Europe.

This was similar to the experiences of early cats and tris, although compressed into a much shorter time period, and with the added pressure of very rapid development of other multihulls making the proas look relatively slow. An exception to this was Crossbow, the 60′ proa which held the world speed sailing title of 31 knots in 1975. Crossbow was very specialised. It only sailed in one direction and had to be towed back after each high speed run. There was also an 18m/60′ long by 24m/80′ wide aluminium proa built for the Rhoute de Rhum singlehanded transatlantic race in 1982 . This boat had a single beam supporting a waterballasted windward hull. The beam could be raked aft to keep the bows up in hard reaching conditions. The maiden voyage was out to the race start, and the skipper inadvertently (possibly with an eye on the gales forecast for the first night) released the rope which held the beam in place. It folded in against the lee hull and the boat capsized.

Cheers has been refitted by some French enthusiasts and is sailing in the South of France. I have no idea what happened to Crossbow, but suspect it is sitting in a shed somewhere in Eastern England. The French monster has, unfortunately, probably been melted down for scrap.

After the ban, interest in proas died out apart from the Pacific Islands where life carried on much as before, although outboard motors replaced sails on a lot of boats. Recently however, there has been an upsurge in interest, mostly people building replicas of traditional proas using modern materials. These tend to be more efficient than the traditional boats, but still have many of their shortcomings. They lack righting moment without large crews, their crabclaw sails tend to be lousy upwind, and because the masts need to be moved from one end to the other, they are slow, cumbersome and dangerous to shunt.

A few people have appreciated and used the reduced stress and materials inherent in proas to build and cruise cheaply. Among them are a couple of French cruising proas sighted in various locations around the tropics, but information on them is hard to come by. Russ Brown (Jim of Searunner trimaran fame’s son) has designed, built, and cruised extensively in Pacific proas, although he is loathe to sell plans or to encourage others, perhaps less skilled, to follow in his footsteps. His current boat, Jzerro spent a winter in Brisbane after being cruised from Seattle via Tahiti and Noumea. It was then shipped back to Seattle.

One of the most travelled proas is in fact Australian. Designed and built by Blaz Kokor in 1980 , About Face cruised the East coast of Australia, before being bought by it’s current owner, Ted Lamont, who cruised, frequently singlehanded from Tasmania to Perth, a journey of some 5,500 miles.

One of the theorists who appreciated the advantages of a proa was Joe Norwood, who wrote about them in his book High Speed Sailing. Norwood appreciated that Atlantic proas placed relatively high stresses on both the masts and the beams, and that these stresses could be reduced by placing these components in the lee hull, as in the Pacific proas. Unfortunately, he confused the issue by adding the complications of hydrofoils which worked in both directions, and the theoretically impressive, but practically very draggy, pyramid rig. None of his proas were built.

What all these proas had in common was that to a greater or lesser degree, they were slow (frequently built to trimaran specs, not appreciating that the loads on a proa can be far lower), and had rigs and steering which made shunting far slower than tacking on a conventional craft. Until these obstacles were overcome, proas were never going to be taken seriously.


I built my first proa in 1992 using stitch and glue ply, an alloy mast and a bewildering variety of steering combinations. At 5m/16′ long, this was competitive with Lasers and slow beach cats, and taught me many lessons about sailing proas. It was followed by U, a 7m/24′ proa with carbon/foam hulls, carbon mast and beams and again, a bewildering variety of steering combinations. Weighing only 150 kgs/330 lbs in sailing trim, U was quick, but was never developed anywhere near it’s potential as I kept experimenting with different ideas. U suffered from 3 different rigs, 3 beam arrangements, 2 windward hulls and a chopped down leeward hull, none of which were ever seriously optimised.


Both the 5m prototype and U used a ballestron rig. This is a rig where the boom is extended forward of the mast and has the jib tacked to it’s forward end and sheeted to a track just in front of the mast (see photos). The forces of the jib and the main balance each other out, resulting in a rig whose angle of incidence is adjusted by one lightly loaded sheet. Known as the AeroRig in the UK and USA and the EasyRig in Australasia, this rig is excellent for any boat, but for a proa is the difference between a 2 minute shunt with lots of hard work and an almost effortless 10 second one.

The steering/leeway reduction conundrum was far harder to solve, but eventually simplicity prevailed, and two balanced rudders were used, of a size which eliminated the need for centreboards. They have very low steering loads, and can be used individually or together.

These boats were both Pacific proas, and sailing them required a fair bit of body movement to keep the windward hull just airborne for maximum speed. This was obviously not the answer for a cruising proa. But then, nor was an Atlantic proa with a highly loaded rig and beams the answer for a lightweight, cheap, easily built boat. The solution had to be to combine the two, by placing the rigs and rudders in the lee hull, and the accommodation in the windward one. The masts could then be stayed to the ends of the beams, thus reducing the bending loads in both, the lee hull could be as long, low and narrow as possible, and the windward hull could be, within limits, whatever size and shape was required to fit the accommodation. Thus evolved the Harry proa, after a lot of sometimes heated discussion with traditional proa aficionados who wanted nothing to do with boats which were ballasted to windward by anything other than crew.

This perfectly logical solution was undoubtedly thought of by many people, but never tried, mainly because of a conviction that the added weight in the windward hull would cause the boat to luff head to wind as soon as it started to move. This has not happened, a trait I attribute to the 2 large rudders, the hull shapes and the balanced rig. Even under main alone, all my boats have been able to shunt and sail successfully.


To test the theory, I built Harry, a 12mx 6.5m proa with full standing headroom, 2 double bunks, galley and toilet in the windward hull, rig and rudders in the leeward one. Built of bending ply * and epoxy/glass, Harry was an incredibly easy boat to build and sail. I could shunt single handed in 8 seconds. It was also fast, easily achieving wind speed, despite, once again, never being optimised. Harry was another experiment, on which I tried various rigs, hulls, beams and rudders. Despite only weighing 650 kgs, it was incredibly tough, withstanding being blown off 500mm high blocks in the boatyard, and later being washed up onto a beach by waves ‘too big to launch a dinghy through’ and pounded for three hours.

Just after I modified Harry to try a single beam layout, we moved to Perth and a trailerable version became necessary to sail in the river and the ocean. Harrigami (Harry + origami) was designed as a folding, trailerable proa with accommodation for up to 4 people for weekend cruising and racing, with the potential to race offshore. It also had to be very cheap and, due to my rudimentary building skills, easy to build.

Mark Stevens and I drew a 15m/50′ harry for Australian, Dutch and Finnish clients. This boat was named Visionarry, for the Dutch boat which was designed to take blind people sailing. These boats were all strip planked timber and had very curvaceous shapes. They required a huge amount of filling, sanding and fairing.

In 2005, we were looking for a better way and hosted 2 KSS workshops with Derek Kelsall. These were a real eye opener about what could be done with foam infused on a flat table. I built a couple of hulls using this method and further developed it to use cheap flat panels moulds which cut the work required and the weight of secondary laminating for joins, fit out, etc.


One of the first clients to use this technique was Steinar Alvestad, a Norwegian building an 20m/65′ harry. Turns out Steinar is a stylist of some repute, so I employed him to redraw all the boats using the new methods.

Why a Harry proa?

Quite simply, proas are the best possible boats for their speed, weight, cost, building time or accommodation. For example Harrigami in sailing trim weighs 550 kgs/1,210 lbs, contains $20,000 worth of materials, was built in a leisurely 450 hours and has room for comfortable weekend cruising with reasonable amenities. Time and cost details are in the accompanying spreadsheets. With 31 sq m/334 sq’ of sail, it easily achieves wind speed up to 16 knots of breeze, despite less than perfect sails, which have been recut numerous times for the various rigs.

Harry proas have the minimum possible amount of structure. Cats and tris both tack. Therefore they see loads from both directions, and have to be built to withstand these. On any given tack, both are carting around a lot of extra boat, solely so that, on the other tack, they will work. Eliminating all the extra bits of boat results in a substantial weight loss. This means that a proa can have a far smaller rig for a given power to weight ratio, which further reduces the loads.


Expeditionarry was designed for a competition run by Wooden Boat Magazine.  The requirement were for a boat suitable for Raids, or expeditions. Fast, overnight, trailerable campers.

The description below is the text and drawings that accompanied the entry.  The text addresses the criteria that the boat had to meet.  There have been some changes to the rudders (liftable), rig (wishbone boom) and folding arrangements (fore and aft) since the boat was drawn. Email Rob for details.


BOA: 6m/20’
Sail Area: 48 sq m/514 sq’
Empty Weight: 360 kgs/794 lbs
Water draft: 100mm-1.75m (4”-5’9”)
Air Draft: 12m/39’8”
Bruce # (empty): 2.5


Expeditionarry is a harryproa. The rig is in the long lee hull, the crew, safety gear, galley, stores and accommodation are in the short ww hull. This allows the crew to stay drier, keeps their weight in the right place and reduces the likelihood of an accidental capsize from uneven weight distribution. 

Except where stated, all the features of this boat have been used and proven on other harryproas built over the last 11 years. In particular, on Elementarry, a 25′ schooner built 6 years ago. For further information on harryproas, see

I have addressed the design criteria individually. This is a very different type of boat, so the explanations are quite detailed. My apologies for this if you are knowledgable about harryproas.

The boat could be built as easily from ply as from composites. I have done the cost/weight numbers for both but the performance calculations are based on the composite version.


Lee hull 10.5m/35′ long x 450mm/18″ waterline beam (23:1 L:B) and a prismatic coefficient of 0.75. This is very high, but proas don’t have to tack, so their hulls can be optimised for speed rather than for turning ability.
Sail area: 48 sq m/514 sq’.
Empty weight: 360 kgs/794 lbs. This may seem light, but Elementarry weighs 120 kgs/265 lbs, the 15m/50′ short handed racer we are currently building is on target for 500 kgs/1,100 lbs and the 50′ cedar strip cruising harryproa at weighs 2 tons. All these weights were accurately predicted, so I am confident 360kgs is achievable.
The “ready to raid” weight (Empty weight plus 160 kgs/350 lbs of crew, 18 kgs/40 lbs of safety gear and 9 kgs/20 lbs of supplies) is 550 kgs/1,204 lbs. The Bruce Number is 2.1. A Tornado with the same load has BN of 1.7 upwind and 2.5 downwind. The T has less righting moment, shorter hulls, a more efficient, though twitchier, rig and no accommodation.
Low windage due to one and a half low, narrow hulls and no standing rigging.
Less water drag than any other multihull of this length. This is a result of the light weight, 2 beam mounted rudders large enough to prevent leeway, and no daggerboards. Wetted surface (excl rudders) of the lee hull with the ww hull flying is 5.5 sq m/59 sq’. Sail area:wetted surface is 8.5:1.
Expeditionarry will perform best with the short hull flying. This is far easier to maintain on a 6m/20’ wide boat than on a beachcat, although it will involve the crew moving to leeward in light air.


The biggest seaworthiness factor for small boats on raid type sailing is crew tiredness, discomfort and fear. To minimise these, Expeditionarry has the following features:

No headsails or extras, which also saves weight, space, complexity, maintenance and cost.
Automatic first reef from the flexible, unstayed rigs,
Simple, stress free shunting, especially compared to a high wind, big wave gybe. Shunting is a little slower than tacking, and a fair bit slower than gybing, but safer and more controlled than either. 
To shunt: Release the mainsheets. When the boat stops, rotate the rudders and pull in the new main sheets as you steer onto the new course. No flogging sheets, travellers, winches, getting in irons, booms hitting heads or crash gybes. The booms are self vanging, so the mainsheet is only altering the angle of attack, not tensioning the leech. Thus the sheet loads are low.
There is no need to steer the boat onto a reach first, as is shown in most shunting sketches. Sheeting in the aft sail first will speed up an upwind shunt (tack), the foresail a downwind shunt (gybe).

Due to the single purchase main sheet (with removable 3:1 fine tune) and rotating unstayed mast, it is very easy to weathercock the rig and safely stop the boat on any point of sail for a rest, a toilet break, to prepare a hot snack or reef, raise and lower the sails.
The lowerable pram hood cuddy on the windward hull keeps the crew dry and out of the wind, but still able to see the sails. The cuddy covers the hatch, so the crew can make hot drinks etc without going below. It also provides standing headroom in the hull.
Ability to anchor or deploy a drogue without going forward on a narrow or open deck. Everything is done from the area between the beams and the hulls.
Ability to single handedly reverse in seconds and sail straight back to a person (or hat) overboard, then stop the boat while it is retrieved.
Comfortable sailing position.

Folding canvas deck chairs to sit in, which can be simply locked in various places on the deck, or stored below when not required. These are much more comfortable than sitting on a flat deck.

For the boat itself to be seaworthy it must:

Heave to without fear of swamping (flush decks, no exposed openings).
Withstand wave capsize when hove to. A 10.5m x 6m (35′ x 20′) platform with 75mm/3″ draft with the rudders raised will withstand very big breaking waves.
Be long enough relative to the rig height to resist nose diving. Mainsail luff length:loa of 1:1 is lower than any other performance multihull that I know of.
Be able to land on a beach in surf (no keels or fixed rudders, sits level when dried out).
Tack and gybe (shunt) in big seas/strong winds
Withstand full speed impacts with the bottom or with floating objects thanks to kick up rudders, no daggerboards and easily replacable 200mm/8” thick foam bows.
Have minimal windage due to unstayed masts and one and a half low, narrow hulls
Be safe if accidentally caught aback. The main sheet leads directly to the windward hull so the sails weathercock with no input from the crew, the boat sits quietly, you reverse the rudders and sail back onto course.
Provide clear visibilty ahead and to leeward, due to no headsails and windward steering position.
Be able to sail off a lee shore in surf using only the sails. This is much easier and safer to do with 2 sails, movable crew weight, raised rudders and occasionally a paddle than it is in conventionally steered boats with daggerboards and rudders. No water foils means it does not get tripped sideways. You go out through the surf with a lot of leeway, but are less likely to be thrown over backwards or sideways by breaking waves.
Be able to be righted after a capsize. This is far simpler than a cat or tri as the masts prevent the boat heeling past 90 degrees. 180 kgs/400 lbs of crew weight standing on the rudders will be enough to right the boat, but getting back on board will require a drop down ladder which is a good idea for other safety reasons as well. 
A less simple option, which doesn’t involve the crew in the water, is to remove the top pins from the windward beam joins and use the ww hull as the righting weight, controlled by the bitter end of the halyard tied to the ww hull.
Falling 6m/20′ onto a sail, boom or mast is not much fun. I would recommend crew tethers be used and either a trampoline with foot sized mesh, or a rope ladder laid across the tramp between the hulls to allow crew to get down from the windward side post capsize.
Pitchpoling a proa is possible, but as long as there is some load in the sails, it will always end up on it’s lee side.


No daggerboards or holes of any sort in the hulls below the water
Unstayed masts
Easily seen and de-weeded rudders. Release a line, the rudder kicks up, then pull it back down.
Crew not required to change sides to tack or gybe (shunt).
The mainsheet and the bendy masts are the primary sail controls, so there are fewer sail shape adjustments than on conventional rigs. The ability to weathercock the sails on any point of sail makes reefing, hoisting and lowering sails much easier than on a stayed rig.

The only deck gear is 2 x 3:1 block and tackles to adjust the 1:1 sheet on each boom. The up/down lines for the rudders are on the beams and the main halyard, downhaul, outhaul and reefing lines are all mast or boom mounted with multiple purchases and cleats.

The booms are mounted on stub masts which rotate in bearings in the hulls. They are self vanging and do not have to be removed to unstep the masts. The sails are stored in the boom with a zip cover.


The portapotti is in the lee hull away from the galley and resting crew and near the centre of pitch. These are important health and sea sickness considerations. If the toilet has to be used while sailing in rough conditions, the boat can be stopped by simply releasing the sheets and will sit very quietly.

There are two bunks 1.9m x 0.5m (6’3″ x 20″) in the windward hull with stowage underneath for food, water, safety gear, etc. One bunk can double as the galley work bench, with the cook sitting on the other, or the stove can sit on the folding shelf on the windward side of the hull in which case it is accessible from the deck, in the shelter of the pram hood. Gimbals are not required, although clamps are.

The pram hood cuddy can also be made to fold all the way over the hard part of the deck, making a 2 person tent and there is room in the lee hull for another bunk. In this case, install a divider for the toilet area.


In my experience 12m x 2.55m (40′ x 8’6″) is a stressful trailering exercise. Therefore I have drawn it 10.5 x 2.4 (35′ x 8′) with the masts 1.5m/5′ over the towing vehicle. The folded boat is 1.98m/6’6″ wide so fits between the wheels on a low cost 2.4m/8′ wide flat bed trailer making it very easy to launch and retrieve. Trailering height is 3.6m/11′ from the road.


Two straps hold the boat on the trailer. The aft one is attached to a beam which also holds the masts. Untie the straps and a bungy over the mast post, insert the gin pole in it’s step in the hull and tie the 3:1 purchase around the 28 kgs/62 lbs mast at or below it’s centre of gravity. Then hoist it to the top of the gin pole. As it goes up, grab the butt of the mast and when it is high enough, align it with the mast hole and lower it. Remove the purchase from the mast and repeat with the other one. This can safely be done by one person in very strong breezes. 

The booms are mounted on stub masts so do not leave the boat. The sails sit inside the booms so only need their covers unzipped. The sail track on the mast lines up with the sail track inside the boom. All that is required is to connect the halyards and raise the sails.

Before doing that, attach one of the halyards to the short hull and pull it tight. Wheel the trailer down the ramp and slide the boat off. As it is flat bottomed, a roller across the back of the trailer and another above the axle are the only rollers required. Neither of these contact the hull when it is fully on the trailer. When the short hull is clear of the cut away on the trailer, it is lowered into the water using the halyard. The beam joins are dynamically unstable at this stage but because the long hull is sitting on the trailer, it should not move. Pull /push the short hull (77 kgs/170 lbs including the beams) away from the long hull until the beams are straight and insert the tapered locking pins in the tubes in the beams and loop the captive elastic over the end. Roll the long hull off the trailer and disconnect the halyard. Load the cooker and anything which cannot be tilted into the windward hull.

The rudders are permanently mounted in rotating housings on the beams. They are trailed in the up position. Release one line, pull on another to lower them, hoist sail and you are away. 

I have successfully used pins to hold folding beams together on larger boats than this, but the two hinge points per beam folding technique is untried in practice. There may be a combination of ramp slope and water depth at which it does not work easily. In which case the long hull could be tied to the trailer to prevent it tipping over, while the short hull is extended and the beams pinned. This would add a few minutes to the launch time. 

From arriving at the ramp to sailing should take two people 15-20 minutes. Retrieving will be a little slower.


The lee hull has watertight bulkheads at the beams, so the ends are sealed buoyancy compartments. There is only one ring frame between the beams, so all this area is available for storage.

Half the volume under the bunks in the windward hull is sealed buoyancy as is the area outboard of the end bulkheads.

The ends of both hulls are glassed over polystyrene which is permanent buoyancy and an excellent buffer in a collision. They are glued on and easily repaired or replaced if damaged.


I do not know any simple, accurate formula to predict this, but the easily managed rig, low windage, high righting moment, shelter on the windward side and lack of things to worry about should make it as easy as possible for the crew. The large, twin rudders and deep reefed schooner rig give excellent, low stress control at low speed with no flogging sails or sheets. Windows in the cuddy allow visibility to windward without exposing the crew. Shunting really comes into it’s own in these conditions, particularly if there are big waves as well.


Either a small outboard mounted on the aft rudder or by sculling oar. The outboard bracket sits on top of the rudder blade and is located by a pin through the out of the water part of the blade. This allows it to be steered with the tiller. A sculling rowlock is mounted on each beam.


The entire boat, (masts, booms, beams, rudders, hulls) is built using flat panels, bent into shape with tailored male/female joins on the edges. Where required, additional laminate is added to the bent shape. For minimum weight, the panels would be vacuum infused foam and glass (plus carbon for the mast and beams) on a full length table, but could also be done on a smaller table. The hulls could also be plywood. 

Either way, there are no building frames required for the hulls or beams (hence no table of offsets) as they are self aligning around the bulkheads. We have built 2 x 15m/50′ infused flat panel proas in the last 12 months and the system works well. 

The panels include all necessary strengthening/stiffening, bulkhead landings, mast and gin pole steps, deck bearings, beam hinges, hatches and hatch rebates, reinforcing for the beams and masts and simple, self aligning male/female joints for the panels. With infusion, it is possible to do all this on the panel, with millimetre (0.04″) accuracy. There is no secondary laminating and all glued components are male/female joints, so it is very simple to build. The panels are all fair and the joins all rebated, so there is no fairing, apart from the 200mm/8″ foam/glass pieces on each bow. Peel ply on all areas to be glued means no sanding is required.

The base laminate is 400 gsm double bias either side of 12mm H80 foam (12 ounce either side of 0.5″ H80 foam), with 100mm/4″ wide strips of unidirectional glass where required and an extra layer of db between the beams. The localised strengthening of the beam sockets and mast bearings is heavy (70k) carbon tow, included in the infusion.

The masts, beams and rudders are uni carbon with glass for the off axis loads. The glass adds a little weight, but saves a lot of money. 

In ply, they would be 10mm/3/8″ lee hull, 6mm/1/4″ windward hull with a couple more frames, stringers along the panel edges and 200 gsm glass cloth on the outside.. The hulls would have smaller deck and chine radii The beams, rig and rudders would still be carbon composite.


Proas do not tack, so hull shapes can be optimised for speed rather than turning ability.

Dory section hulls do not need floors which is a saving in overall height and weight, are easier to get on and off the trailer, have better resistance to loads when sitting on the beach and are easier to build.

The hulls are identical , except the stations are closer together by a factor of 20/35ths on the windward hull.

According to Michlet (drag program), the chosen hull shapes are superior to conventional semi circular shapes at speeds above 5 knots, slightly inferior below, due to the slight increase in wetted surface. This is more than offset by the low weight of the boat and higher top speed. 

Note: Rick Willoughby performs my Michlet calculations. Among other boats, he has designed the hull of the human powered 24 hour world record holder. Rick designed, built and tested a half size Expeditionarry hull which performed as predicted by Michlet.


The rudders are mounted in robust housings which can rotate around the beams. They are oversize so no daggerboards are required. Blades are NACA 0012 sections with substantial integral carbon shafts. Fore and aft lines hold them down and include simple wooden dowel fuses to allow the rudders to kick up in a collision or grounding. They are easily lifted to clear weed, sail in shallow water, run up the beach or to remove tripping forces when hove to in big waves. The front one can be raised when not sailing upwind.

They rotate through 360 degrees which makes shunting (and sailing backwards) much easier.

The tillers are pinned through the top of the shafts and can rotate vertically. This places both tillers over the trampoline while sailing, meaning both are pushed to luff, pulled to bear away and in the event of the rudders kicking up in a grounding/collision, the tillers are not in the water, where they could be broken.


Composite materials for hulls, decks, beams, masts, booms, rudders and rudder mountings plus 10% wastage: $8,944
Hardware, rope, paint, glue, mast and rudder bearings, beam pins, portapotti, cooker: $1,953
Sails and battens: $5,000 (range from $3 -$6,000)
Total for home builder: $15,897, plus table and consumables
Labour and consumables if a pro built the boat: $28,800. The first one would probably cost a bit more. After the third one, they’d be a bit less.
Total for pro built: $44,697
Costs are Australian bulk (rolls of cloth, drums of resin). 

The laminates have all been engineered by Etamax Engineering.

The spread sheet areas and drawings are from Rhino.

The costs agree with the numbers from the 2 x 50′ infused flat panel harryproas we have built. These are more commodious than Expeditionarry and sell for $35,000 plus rig, so $45,000 for a raid type 35′ folder ready to sail is about right. 
An inexperienced infuser would take at least twice as long as we do and also has to build a table, but a complete novice could assemble pro built panels with ease.

The pram hood, trailer (approx $2,000), motor, sail covers, sculling oar/outboard and safety gear are not included.

The okoume ply (see bottom of spread sheet) hulls are heavier (444 lbs vs 370) with cheaper hull materials ($2,422 vs $3,183) and more labour. Rudders and rigs would still be composite, beams could be timber and ply with similar tradeoffs to those for the hulls.


“Raids” vary from 12 hour+ race legs with both crew trapezing (Worrell etc) to Puddle Duck Racers cruising in company (Texas 200). I have drawn a boat biased towards the former, but there are a number of options, depending on how serious the crew are about performance.
a) The chosen rig is a schooner to fit within the length requirement and to make it easy to sail for a novice. A race oriented boat would be wing masted and might have a una rig. 
b) The hard deck could be smaller, slatted or partially replaced with a trampoline.
c) The lee hull could be longer or shorter simply by moving the stations. The windward hull could be scaled up 150% without problems, although as the weight increased, the scantlings would need to as well.
d) If the boat is going to be abused (rocky beaches, etc), then high density foam or infused plywood (a very tough material) could be used in the hull bottoms. 

Harryproas are very different to other boats. If any of the above is unclear, unbelievable or questionable, please do not hesitate to contact me at

Rob Denney


Rob Denney is a New Zealander who started sailing at age 4 and has scarcely been out of a boat since. He has sailed in 7 Sydney Hobarts (2nd on handicap, fastest elapsed time), 6 Brisbane Gladstones (1st on handicap, fastest elapsed time, race record) , worked as a professional yacht skipper, raced and cruised extensively and designed and built some very radical craft, among them a windmill powered catamaran, a precursor of the Formula 40 class and an ultra light 12m catamaran with independently pitching hulls, variable beam and unstayed rig. He owned a successful boat building materials business, was a pioneer in low cost carbon mast building and is an occasional yachting journalist. He built his first proa in 1995 and has been building, sailing and experimenting with them ever since.

Rob and Steinar enjoy talking boats, building and sailing. Please email them on


to discuss your requirements or your project, whether it is an off the shelf or custom design, or an idea which you think has potential.