Heraeus Noblelight Limited v First Light Lamps Limited

Patent claims are construed purposively, but deliberate words in a claim retain meaning. A numerical or physical limit may have a fuzzy boundary where a small deviation has no material effect on how the invention works and the skilled person can identify the practical limit. A method using a bead marginally larger than the claimed external diameter therefore infringed where the excess did not impair seal strength or utility.

Infringement by equivalents is assessed using the Actavis questions. A de minimis principle applies where alleged infringement results only from occasional inadvertent departures from the claimed method. Obviousness must be assessed without hindsight and separately against each item of prior art.

Heraeus alleged that First Light’s manual and partly automated methods for sealing quartz arc-lamp tubes infringed EP (UK) 1 598 845. The patent claimed direct sealing using a sealing-glass bead larger than the tube’s internal diameter but no bigger than its external diameter.

First Light used beads stated to be 0.1mm to 0.3mm larger than the external diameter. It denied infringement and counterclaimed for revocation for obviousness over the Mathijssen and Churchley patents. The issues were the construction of the bead-size limitation, infringement on purposive construction, by equivalents and literally, and obviousness.

1. Infringement on purposive construction. The words “not bigger than” did not impose a rigid measurement limit. Read in the context of the patent, they included beads not materially bigger than the tube’s external diameter, where the excess could have no material effect on how the invention worked. The relevant practical limit was whether the excess impaired the strength or utility of the seal. Beads within common general knowledge tolerances were therefore within the claim.
2. The Annex C method, and the materially identical Annex D method, infringed on that construction. The patent’s technical purpose was to address weakness in the seal, not to achieve exact matching of external dimensions or satisfy every possible customer tolerance.
3. Equivalents. Applying the three questions in Actavis, the variant achieved substantially the same result in substantially the same way. The same reasoning applied to the question whether strict compliance with the literal wording was intended. The equivalents case therefore succeeded.
4. Literal infringement. Inadvertent departures from the Annex C process would infringe only if they occurred on a more than de minimis basis. The evidence did not establish either that the sample used an undersized bead or that such errors occurred more than de minimis.
5. Validity. The patent was not obvious over Mathijssen or Churchley. Mathijssen taught a specific stress-management geometry which the skilled person would not discard without a fair expectation of success. Churchley likewise retained an indirect-seal structure and did not make removal of the GS10 annulus or direct sealing obvious. The patent was therefore valid.