This is just a FYI in case anyone is interested. A problem solved, but a mystery not solved. On and off through 2019 there were discussions about how to predict or debug scrap stacking order. This message by Tarquin covers the expected behaviour fairly comprehensively. https://www.mail-archive.com/therion@speleo.sk/msg07925.html That is all very well, and 99% of the time my projects stack scraps the way I expect. Except that I have one large project where lower passage scraps stack on top of upper level scraps, in a few locations only, in the scenario where no maps defined whatsoever (with the intention that scraps therefore stack according to average altitude). Other scenarios using the same source data, where maps are defined and selected, all plot as expected according to the map structures. But for large overview maps I need to plot all scraps without any map structure, and I'd like them to stack correctly. The offending scraps were drawn many years ago, and were all much longer and encompassing much greater passage height variations than I would adopt these days. I just assumed that the average heights that Therion calculated were somehow not collated in the stacking order that I expected. Anyway I eventually found the time to check the scrap and heights reported in therion.log. To my slight surprise I found that in fact the average heights that Therion calculated were in exactly the order I expected. So why was Therion not stacking the scraps in the expected order? My anticipated solution of breaking the scraps into pieces to make them behave properly no longer seemed likely to make any difference. Since it was implemented, I have been making use of the 'maps off' to disable previews and offsets, as a more refined way of achieving a particular output than just 'not defining' any maps in a particular thconfig. https://www.mail-archive.com/therion@speleo.sk/msg07571.html In desperation I thought I'd try adding 'maps off' to the thconfig above, where no maps were defined. It should, I expected, make no difference at all. And in one easy step, all of the scraps now plot in the correct stacking order! So why does 'maps off' make scraps plot in the correct order when there are no maps defined? Bruce