Technically speaking, the problem seems to have been a lack of support under the frame. This manifested itself in a twisted hinge at the top right corner, and the final leaf that would not fit properly into the frame, so the paintwork of the frames has been damaged.
"We have no problems, only solutions," they announced.
Using a spirit level they estimated that the frame was off the level by about 1mm in height per metre of width, or three millimetres over the whole width of the window. This was perplexing as the frame was set to within a millimetre, using a laser.
Adjustments to the screws on the hinges did not work, and they removed the rotated hinge at the top right corner, then replaced it, but it just went back to its original angle.
The hinge will only support up to 130 kg, so most of the weight of the windows should be supported by the rail at the bottom. Each leaf weighs around 80 kg, so two will weigh 160 and three 240. The weight is all transferred onto one spot on the rail, and when the window was installed, plastic spacers should have been used at regular intervals to stop the frame and rail from sagging. It seems that wood was used under the frame, which had disintegrated through the weight and through rain getting in from outside. When the window was installed, there was clearly nobody there who knew all this.
The people from Pazen came to this conclusion by removing the caulking around the bottom, which we had put there, probably unnecessarily, worrying at the lack of airtightness from the Compriband. They fixed the problem of the sagging rail by inserting small squares of hardwood, individually manufactured by our master carpenter from left-over bits of floorboard. They had hoped for some plastic spacers, but there were none in the house or in the importer's stock. As the outside of the window frames are now sealed, moisture should not be a problem, and the wood should last for a while. There is now no caulking along the bottom of the window frame, but it looks like it will need to be put in again, perhaps when the tiles are put in.
In the process, I think the Germans learnt the English word for "crow bar" and I learnt the Japanese word, which is disappointingly obvious: ba-ru. They needed this to lift the frame up and insert the wooden spacers.
Of course, problems have human dimensions as well as technical dimensions, and in human terms, the basic problem has been that information on installing the windows did not get to the people actually installing them.
The language barrier could be blamed for this, but I think the biggest problem has been incompatibility between our architect and the window importers. The importers, perhaps justifiably, see the details of window installation to be the responsibility of the architect. The architect, totally unjustifiably, probably sees these foreign windows and the people trying to bring them in as unnecessary, and has been looking for problems with them from the beginning.
As far as Pazen is concerned, they think they sent a high-quality product, which has been compromised in its installation. They were assured by the importers that they knew all about windows, which may be true, but the information did not get to the people putting them in. These windows were perhaps five times heavier than the windows they were used to, were installed from the inside rather than the outside and had unfamiliar mechanisms, so clearly important information had to be communicated.
Pazen came last November, coinciding with the planned installation of the windows working back from the January NEDO deadline, and gave various instructions and answered several questions from the architect. Two of the instructions, which the architect carefully noted in his book, were that the screws should be spaced 15cm from the top and bottom of each side, and with less than 60 cm between them. In other words, a window with a side 90 cm or less only needed two screws. This message did not get through to the carpenter installing the windows, who put screws in every 50 cm, and then ran out, and is still waiting for more screws two months later.
Also, the screws should go, not through the middle of the frame, but near the inside edge, as the boss clearly told the architect last November, and the architect clearly wrote in his note book. At least it was clear that he wrote it, although we cannot clarify whether he wrote it clearly. You can see from the picture above where the screws went, and looking at the cross section picture of the window on the left, that's going straight through the middle of the frame, where there is insulation, and it is not the best place structurally. The screws should have gone where the carpenter's thumb is.
So it seems that the architect and importers have been negligent in guiding the builders in the installation, and I wish that instead of worrying about my job, I had made more effort to work between them. I should have made sure I was there at the meeting between importers and builders, and I should have made sure the builders were there to meet Pazen when they came, instead of being left out of the loop.
Another question I have, though, is whether Pazen's design is right for this size of window, or whether the three leaves are just too big and heavy with triple pains. As I was looking at the great size and weight weighing on the little hinge, I was wondering whether it wouldn't have been an idea to have another hinge on the right hand side.
Or perhaps it was just me being stupid in wanting this window. If this, and the other windows on the South wall, had been double glazed rather than triple glazed, they would have been substantially lighter, and this may not have been such a problem. They would also have been cheaper, and by my calculations would not have resulted in any net heat loss as double glazing lets in more light than triple. Triple windows will certainly be nicer in the winter when it's really cold outside and we are right next to them.