I received some feedback from GBCI that is specific to a project, but that I think will be broadly applicable and may help to answer some of the questions below. GBCI realizes that the Reference Guide includes content that is a copy/paste from the energy M&V section, some of which isn't the most appropriate or applicable to water.

The comments are long, so I will provide a synopsis first: You have two options - you can either use the first year of data as the baseline (more common) or you can attempt to calculate a baseline based on design data (more challenging). There is not currently a protocol in place for the latter, though ASHRAE has a committee working on this; you would need to create your own and explain it to GBCI. If you choose this, you will have more requirements in the M&V plan - the more complicated your approach, the more documentation you will have.

The remainder of this post is an excerpt from GBCI's comments:

Some major differences between WEc2 and EAc5 include the following:
• The HC ref guide states on p. 93 that the “baseline for the calculations is based on one year of actual measurements” in regards to Option B. If teams choose to estimate/model/calculate the water consumption to establish the baseline condition that would still be acceptable and would be more consistent with the IPMVP; however, that is not required per the reference guide.
• The water end-uses for which the M&V plan will be provided will likely be submetered as submeters are specifically required for most of the water end-uses (there is not such a prescriptive energy metering requirement within EAc5).
• The HC rating system does not have quantified water savings credits associated with many of the end uses unlike EAp2 does regarding energy savings
• As the team has pointed out, the tools, techniques, and standards that define water modeling haven’t been developed, at least not to the same extent as they have for energy modeling (e.g., ASHRAE Standard 191 is just now under development and is looking to address some of these shortcomings).

So a team may choose to utilize the first year of consumption as the baseline case, the credit becomes less of concerned with M&V of the “savings”, but rather a plan to compare performance against that of the first year of operation and demonstrate sustained performance. This is not really consistent with Option B or Option D of the IPMVP but based on the way this is written in the reference guide we must allow it. But given the challenges in establishing an appropriate baseline and readily and reliably calculating the projected baseline water consumption, it is understood why this option was added to the requirements.

Alternatively, if a project team wants to establish a baseline for a water end-use rather than utilize the first year of operating data (note neither LEED nor the IPMVP puts limitations on what the baseline must be), this would follow more closely to the Option B path described within the IPMVP. The projected baseline water use is determined by calculated in the hypothetical water performance of the baseline system under post-construction operating conditions, and the savings are determined by subtracting the measured end-use consumption from the projected baseline.

In most cases the water conservation measures and water consumption of various end uses are expected to be readily isolated, thus allowing for Option B to be utilized (or the modified Option B utilizing the first year of post occupancy water consumption as the baseline). The opposite it generally true for EAc5 because the energy conservation measures typically implemented within most new construction projects are too highly interactive to allow for verification of savings with any method other than Option B. I can see Option D Calibrated Simulation being valuable in some situations if you have complex systems such as cooling towers, if one had a specific baseline condition from which savings needed to be verified, if monitoring plans were in place for appropriate calibration parameters like weather, operating hours, and cooling system loads, and if a software program was available such that one could calculate consumption within the desired level of accuracy. But as noted above, we would also allow the first year of water performance to be considered the baseline even for systems like cooling towers.

As far as the M&V plan is concerned, our expectations are that the plan contains the items listed within Section 3.3 M&V Plan of the IPMVP Vol I Energy and Water Savings 2002 that can be reasonably applied to water and the method for which the Baseline has been established. The more complicated the baseline, the more thoroughly it needs to be documented within the M&V plan.