A pilot forum discussion out of Livermore, California (KLVK) has surfaced a procedural ambiguity that touches on a recurring headache for instrument-rated pilots: departure procedures whose depicted flight paths become geometrically unachievable for aircraft with any reasonable climb performance. The LVK3 departure instructs pilots to fly heading 255 until 1,200 feet, then execute a climbing right turn to heading 020, intercepting the OAK VOR/DME radial toward the ALTAM fix. The problem, as the original poster identifies, is straightforward trigonometry: any aircraft with typical climb performance that turns to 020 at 1,200 feet will end up east of ALTAM rather than achieving the intended intercept of the OAK 060 radial from the west. This is compounded by a published Letter to Airmen (LTA) explicitly directing pilots to fly the conventional procedure exactly as depicted, even when using RNAV equipment, rather than allowing a direct-to substitution — despite the fact that flying it as depicted appears to produce a track that doesn't actually accomplish the procedure's stated purpose.
This is not a hypothetical curiosity. The poster describes watching multiple aircraft ahead of them simply fly direct to ALTAM in the turn, deviating from the literal depicted procedure, without any correction or comment from NorCal Approach. That real-world controller tolerance is telling: it suggests a gap between the letter of the published procedure and how ATC actually expects (or at least accepts) it to be flown in practice. For working pilots, this is the kind of institutional inconsistency that creates real liability exposure. If a controller on a different day, or a check airman during a proficiency check, holds a pilot to the LTA's literal language — fly the conventional procedure exactly as depicted — while the geometry of the procedure makes that technically impossible without ending up in an unintended position, the pilot is caught between two irreconcilable directives. The reference to a nearly identical four-year-old thread about the OAK ILS 28R missed approach, where the same geometric disconnect exists at Oakland, underscores that this isn't a one-off drafting error but a pattern in how legacy conventional (non-RNAV) procedures get published and retained even as aircraft performance and area navigation capability have moved well past the assumptions baked into the original procedure design.
The broader issue here is one that affects instrument pilots at busy Class C and D airports throughout the National Airspace System: the slow, uneven transition from ground-based conventional navigation procedures to RNAV-based design. Many SIDs, STARs, and missed approaches at legacy airports were built around VOR/DME geometry and climb gradients calculated for aircraft performance profiles that no longer reflect the bulk of the fleet, particularly turboprops, light jets, and even well-performing piston singles and twins climbing steeply out of a 1,200-foot AGL turn altitude. When RNAV capability is layered on top of these older procedures — as it increasingly is, since most GA and business aircraft now fly with WAAS-capable GPS — you get situations where the "as depicted" conventional procedure and the "direct-to-fix" RNAV solution diverge, and controllers, TERPS specialists, and local FSDO guidance don't always agree on which one governs. This is precisely the kind of ambiguity the FAA's broader PBN (Performance Based Navigation) initiative is meant to resolve over time, but until individual procedures like LVK3 get formally redesigned or amended, pilots are left making judgment calls in real time based on what "seems to work" rather than clear written guidance.
For flight departments, training programs, and individual pilots operating into airports with similarly dated conventional departures, the practical takeaway is to treat published LTAs and procedure notes with real scrutiny before departure, not as boilerplate. When a depicted heading-and-altitude sequence produces a track that cannot geometrically achieve the stated navigational intercept, that discrepancy should be briefed, documented, and ideally raised with the FSDO or through the FAA's Aeronautical Charting Forum process rather than resolved silently by mimicking whatever the preceding traffic did. Controller silence on a given day is not regulatory sanction, and pilots who assume otherwise may find themselves without cover if an incident, deviation, or enforcement action ever turns on the letter of a procedure that was, by the pilots' own admission, never actually flyable as published.