Shade is architecture, not decoration

The first drawing I made of this estate was not a plan. It was a stack of eleven tracing sheets, one for every daylight hour, each with the shadows inked in where they fell on 21 June. I laid them over each other on a light box and the building more or less designed itself out of the gaps.

That sounds like a nice story and it is also literally what happened. We had nine hectares, forty-one mature palms and a client brief that amounted to: it must be bearable in August without machinery. In this climate that is not a sustainability position, it is an engineering constraint. Air conditioning here would have cost more to install than the roof and more to run than the staff.

Shade is not one thing

The mistake people make is treating shade as a binary — a place is in it or out of it. It is not. A palm canopy gives dappled shade that drops the air temperature by perhaps three degrees and the surface temperature of anything under it by fifteen. A solid roof gives deep shade, cooler still, but it also stops the air moving if you build the walls wrong. A wall gives you hard shade that is cool for an hour and then gone.

The east arcade is the clearest example. It takes the low sun straight on until about nine, and then the row of palms that were already there takes over for the rest of the day. We could have roofed it and had deep shade from dawn. We did not, because that hour of low light is the reason people sit there for two hours over breakfast instead of twenty minutes.

We spent the first summer doing nothing but recording where the shadow fell. The plan came out of that survey, not the other way round.

Building to the survey

Two of the five buildings sit on old farm footings, so their position was decided in about 1890 by someone who understood this ground better than we do. The three new ones were placed on the tracing sheets first and on the site second.

Cisterna went over the old water tank because a metre of stone and a void underneath is the best passive cooling available anywhere on the property. Palmar went into the densest part of the canopy and is consistently the coolest set of rooms we have — guests notice within about a minute of walking in, and we have measured the difference at four degrees in the afternoon.

Mirador is the exception and the argument we had most. It faces north-west, it takes the full afternoon, and by the survey it should not exist. It has the view. We built it anyway, gave it the thickest shutters on the estate and told the truth about it in the room description, which is why the people who book it are never surprised.

What we would do differently

Plant earlier. The nine palms we put in during the first spring are only now, fifteen years on, throwing usable shade. If we were starting again the planting would go in three years before the first foundation, and we would site the west terrace two metres further back to catch them sooner.

The other thing is that we drew for June and should have drawn for September. June is the extreme, but September is when the estate is fullest and the sun is lower and further south. The west terrace loses its shade about forty minutes earlier than the June survey implies, and for two weeks in late September the long table gets a strip of sun across one end that nobody wants to sit at. We move the table. It is a small failure but it is the kind you only find by living in the building.

Everything else has held. Fifteen summers, no mechanical cooling, and the rooms sit at twenty-four degrees when it is thirty-six outside. The palms did that, mostly. We just read them carefully and tried not to get in the way.