When a valve is manufactured, it undergoes a process called evacuation where successive pumping operations remove air from the envelope prior to sealing. However, even after evacuation, a small amount of gas remains - one part per one-hundred-thousand-million - as Cilla would agree 'That's NORRA lorra, lorra gas!' More gas may be released from the metallic components that comprise the electrode cage of a valve, gas that is trapped within the interstices of the metal structure especially if a valve is run under arduous operating conditions such that it gets extremely hot.
Valve manufacturers include a component within the valve that 'mops up' residual gas - this is the getter flashing - that pretty silver mirror like coating seen on the inside of a valve's glass envelope. The flashing is a deposited barium metal mirror, which originates from the getter rings and pans that form part of a valve's structure. After evacuation the sealed valve is subject to an inductive RF heating pulse which evaporates the metallic barium from the getter ring/pan onto the envelope.
This mirror is sacrificially consumed as it 'mops up' gas within the envelope and many of us will have seen an elderly valve where the getter flashing takes on a brown and then a pinky-white hue as it's absorbtive properties have exhausted - all schoolboys of a certain age will also be familiar with the silver getter flashing disappearing to a dull white powdery coating as the valve envelope is broken either by stones, airgun pellets, Dinky toy collisions or other means thus letting atmospheric oxygen in to consume the getter flashing!
OK, so we have got a bit of gas left in a valve and something to sort out any that's left, what does the gas actually DO, can it be a problem????
The answer is a resounding YES for as the gas molecules 'bimble' and 'gambol' in Brownian motion within the envelope, electrons from the cathode 'belt' into the gas molecules causing them to shed their own electrons and become positively charged cations. These positively charged gas molecules are then attracted to the negatively charged grid where they tend to nullify the grid bias making the anode current increase. Unfortunately this can start a cascade effect as mure current causes more gas molecules to ionise and the anode current can continue to rise such that a sort of meltdown occurs in which the valve destroys itself.
In reality though. often long before this happens, the positively charged gas molecules are simultaneously attracted to the cathode where they bombard the emissive coating, forming a non emissive layer on the cathode surface called 'poisoning' which reduces emission to very low evels and renders the valve useless. Could that be what has happened to that bargain valve with 5% emission - is it a victim of gas attack?