Dimming Lights - Internal Lighting Power Allowance

Unless you have an automatic control, you need to assume maximum lighting power. I'm confused by your statement - "Do we assume that the light bulbs are always dimmed to the max (and therefore use maximum power)?" - dimming generally results in energy savings. "Dimmed to the max" is better known as "off". If you are following the ASHRAE 90.1-2007 prescriptive path for compliance, you cannot take any credit for dimming by the BMS. If you are modeling with Appendix G, you can model and take credit for any daylight responsive controls, including dimming. Other automatic dimming does not fit easily into the additional controls outlined in Appendix G, which are allowed to use the Table G3.2 power adjustments. However, If the dimming is time-clock based, controlled by the BMS, I think you could submit that as an ECM. You should model the lighting under the BMS control sequence, but you should also be prepared to explain to the LEED reviewers why the modeled levels are appropriate, and account for local overrides (if possible) in the modeling.

I also have same kind of question: I am modeling a building using Appendix G. One of the strategies is to implement daylighting sensors with a continous dimming control. So the main question is: What is the minimum output fraction allowed? For the moment I am assuming that the lighting power can be reduced up to 10% of its nominal capacity. Is this ok? Many thanks in advance!

The minimum output fraction is typically dependent on the dimming ballasts. Some dim to 0%, most dim to 10%. Check the specified/installed ballasts to see what you should use in the model.

Either the specific parameters must be known or the tenant guidelines must contain the specific parameters. You can only claim savings for strategies that have been designed or are sufficiently specified in the tenant guidelines to enable them to be modeled.