Summary
Shared demand = Σ over uses of [base ratio × size × month factor × day factor × hour % × modal adj × non-captive %], evaluated for every hour/day-type/month; governing condition = the max cell. Reduction = governing shared peak vs Σ individual peaks.
Build order (spreadsheet)
- Program table: each use, size, base ratio (ULI or accepted local/ITE substitute — cite which per use).
- Pull ULI time-of-day %, monthly, weekday/weekend factors per use — record table numbers.
- Adjustments row per use: modal (transit/walk context — justify), non-captive % (share arriving for that use vs already on site), internal-capture interplay documented (don’t double-discount).
- Matrix: uses × 24 hrs, one sheet per day-type × critical months (or all 12) → hourly totals.
- Governing peak: value, hour, day-type, month. Sanity-check: does Dec-retail / weekday-office logic match intuition?
- Compare: governing shared peak vs supply vs code vs Σ individual peaks.
Worked micro-example (illustrative)
Office 40k SF + Restaurant 6k SF: office peaks ~10am–2pm weekday, restaurant ~7–9pm & weekend — shared peak lands early-afternoon weekday or Fri eve depending on ratios; a ~10–25% reduction vs sum-of-peaks is a common outcome when factors are defensible (never promise a number before the model runs).
Hand-check
Recompute one governing-hour cell per use with a calculator; matches sheet exactly.
Example application
Attach the workbook to the project’s shared-parking note; assumptions table into report appendix verbatim.
Cross-links: Workflow - Parking Study · ITE Parking Generation · Parking Code Comparison
Practice question
Why can a 100%-captive hotel-restaurant add ~zero demand, and what evidence supports a captive claim?
QA/QC reminder
Every factor cell traceable to a ULI table number or a written justification — no naked percentages.