Well, a little google-fu reveals this: http://en.allexperts.com/q/General-History-674/Sewers.htm
First, the Romans did not invent sewers. Many ancient civilizations, including the Babylonians had them. There have even been some evidence of small sewer systems in Scotland dating back to 3000 BC. ...
An interesting example in this discussion is
Mohenjo-daro, which was originally built in 2600 BC and abandoned around 1500 BC. It was destroyed and rebuilt about seven times over those 1,100 years; it is believed mostly due to major flooding. (In general terms, the city's structures and even foundation weren't reliably proof against
100-year floods.) Some selected quotes to spark thought:
Mohenjo-daro is located in Sindh, Pakistan on a Pleistocene ridge in the middle of the flood plain of the Indus River Valley. The ridge is now buried by the flooding of the plains, but was prominent during the time of the Indus Valley Civilization. The ridge allowed the city to stand above the surrounding plain. The site occupies a central position between the Indus River valley on the west and the Ghaggar-Hakra river on the east. The Indus still flows to the east of the site, but the Ghaggar-Hakra riverbed is now dry.
...
Anthropogenic construction over the years was precipitated by the need for more room. The ridge was expanded via giant mud brick platforms. Ultimately, the settlement grew to such proportions that some buildings reached 12 meters above the level of the modern plain, and therefore much higher than this above the ancient plain.
So one of the most successful large early cities planetwide was at the intersection of a huge river trade network (the Indus valley), and another river system, situated on a geographically prominent ridge. The more I research, the more I'm convinced that the basic recipe for a successful early major city is to place it on (comparatively) high ground along a major river, around or near some other significant trade feature... another river, the ocean, an overland caravan route, etc.
Mohenjo-daro has a planned layout based on a street-grid of rectilinear buildings. Most are of fired and mortared brick; some incorporate sun dried mud-brick and wooden superstructures. The sheer size of the city, and its provision of public buildings and facilities, suggests high levels of social organisation. At its peak of development, Mohenjo-Daro could have housed around 35,000 residents.
The city had a central marketplace, with a large central well. Individual households or groups of households obtained their water from smaller wells. Waste water was channeled to covered drains that lined the major streets. Some houses, presumably those of wealthier inhabitants, include rooms that appear to have been set aside for bathing, and one building had an underground furnace (hypocaust), possibly for heated bathing. Most house have inner courtyards, with doors that opened onto side-lanes. Some buildings were two-storeyed.
...
Close to the "Great Granary" is a large and elaborate public bath, sometimes called the Great Bath. From a colonnaded courtyard, steps lead down to the brick-built pool, which was waterproofed by a lining of bitumen. The pool is large – 12m long, 7m wide and 2.4m deep. It may have been used for religious purification.
...
Mohenjo-daro had no circuit of city walls but was otherwise well fortified, with towers to the west of the main settlement, and defensive fortifications to the south.
...
The city is divided into two parts, the so-called Citadel and the Lower City. Most of the Lower City is yet to be uncovered, but the Citadel is known to have the public bath, a large residential structure designed to house 5,000 citizens and two large assembly halls.
In general, Mohenjo-daro had water and sanitation infrastructure that most folks would not have expected until the
sanitation of the Roman Empire; by comparison, Rome had the first simple open sewers between 800 BC and 735 BC, more major work around 600 BC, the main
Cloaca Maxima not enclosed until 33 BC, and the system generally in full form by around 100 AD. A possibly relevant quote for those looking for early "adventurer-compatible sewers", from
Strabo's Geographica (info dates from the ~20 BC to ~23 AD range):
The sewers, covered with a vault of tightly fitted stones, have room in some places for hay wagons to drive through them.
The main failing in the overuse of the
Absurdly Spacious Sewer trope isn't that they didn't exist (see the extensive Real Life section on that page for some more examples), it's that they were historically a rare feature of the world's major cities only. Too many video games put a Paris-scale catacombs + sewer under some tiny starting town.