We have a 2000 page city info website and for the last year have been building data such as hotels, leisure facilities, events etc, all geo coded. We can now produce unique auto generated pages of information for every postcode in the area. On checking, it means adding twenty fold the number of existing pages and I am concerned we could incur a Google penalty for such a high instant increase in page volume.
My question is:
1. Do we noindex,nofollow and filter them in gradually ie daily
Let me throw in a tiny bit of factual information, even if it goes in the "file and forget" bin. About a month back I added a slew of auto-generated pages in one go. The numbers don't match yours, but it works out to about 1/4 the site's total pagecount. They're all individually marked "noindex" (on the rock-solid grounds that they have no unique text content and I don't want their images indexed).
When the googlebot got around to crawling them, there was a one-day hiccup in the "pages indexed" graph on gwt and then it went back to its previous number. This suggests that indexing is not two stages but three: first the crawl, then the crude index, and finally the processing of the index including filtering-out "noindex" pages. I guess it's really four stages if you add in any final penalties that happen on the fly in the SERP-display phase. But noindexed pages would never get to this stage.