Lighting Knowledge

Should ballasts and capacitors be bypassed when installing LED tubes?

As a company we have been performing these types of energy efficiency upgrades first with T5 adaptors and now with LED tubes since 2006.

The issue of power factor in fluorescent luminaires, indeed power factor in general, seems to be a much misunderstood or ill-informed area even among electrical engineers and master electricians.

As a consequence there is much “FUD” surrounding the topic. Fear, Uncertainty and Doubt. Common beliefs are:

  1. Ballasts left in circuit still use power and reduce the energy saving

  2. Ballasts left in circuit are a point of future failure

  3. Leaving capacitors in circuit with an LED tube creates a “bad” power factor

  4. Leaving capacitors in circuit creates a leading power factor

  5. Leaving capacitors in circuit increases current

  6. If the current increases so does the power and so there is no energy saving

  7. If current on the lighting circuit increases the Volt/Amp (VA) demand charges from the retailer will increase

It is arguable that there are few companies in the country with more experience and knowledge of the power saving and electrical aspects of such installations. So let me give you a different point of view.

  1. Ballasts left in circuit still use power and reduce the energy saving

    For this discussion we will use a 36W fluorescent tube and a B2 type magnetic ballast as they are commonplace in the market place. This configuration consumes a total of 42-46W. Leaving the ballast in place and connected consumes about 0.5 watts per hour, we know this from bench testing. When using a 20W LED tube the savings are from 22 – 26 watts so it is insignificant.  Using operating hours of 3000 hours per annum and 0.25 cents p/Kwh the cost of removing control gear in order to save the 0.5 watts per hour will save 50 cents a year and cost about $20. The Return on Investment being 40 years, this is seldom a motivating reason for the average accountant


  1. Ballasts left in circuit are a point of future failure

    1. The purpose of a ballast is to “choke” or limit the current getting to the fluorescent tube. Think of it as a flow restrictor on a hose. During this process the ballast produces significant heat.  Without it the Fluorescent tube would blow its cathodes which we know from bench testing without a ballast in circuit (don’t try this at home!).  When a fluorescent tube of 36 watts is replaced with an LED tube of 20 watts, the current required by the tube never reaches the level where the ballast is required to start choking. If the ballast doesn’t have to work it doesn’t generate heat and has effectively become a bit of wire which can remain in the light fitting indefinitely as the effects of heat have been removed. 

  1. Leaving capacitors in circuit with an LED tube creates a “bad” power factor

    1. “Good” power factor is unity or close too.  “Bad” power factor is considered less than 0.8. Ballasts are inductors and have a “bad” power factor of about 0.40 (lagging). Capacitors are the opposite of inductors and are installed to counteract the inductance of the ballast and so correct the PF to better than 0.8 lagging.  LED tubes are capacitive so installing them on luminaires with fully functioning capacitors will cause the luminaire to move from an inductive (lagging) PF of typically 0.8 to a capacitive (leading) power factor of approximately 0.4.

    2. In isolation the statement is correct. Leaving fully functional capacitors in circuit with an LED tube creates a “bad” power factor”. However the statement should not be taken in isolation as will be expanded on below. A leading power factor in the ceiling may be very beneficial to the overall site energy supply. See point 7e below.


  2. Leaving capacitors in circuit creates a leading power factor

    1. As stated above this is true.  For some reason leading power factor is sometimes viewed as a negative but as explained above it is just the opposite of lagging.  No doubt this has come about because most electrical loads are inductive and electricians and electrical engineers have come to believe that power factor is always lagging. The amount of leading or lagging power factor is important but the direction doesn’t matter and can indeed be beneficial. A leading power factor in the ceiling may be very beneficial to the overall energy supply. See point 7e below.


  3. Leaving capacitors in circuit increases current

    1. The reason it is desirable to have power factor close to unity is because at unity the current used is proportional to the watts.  The further away from unity the PF gets then more current is needed to produce the same watts.  It is called reactive current and is effectively wasted.  The energy provider has to provide more amps to produce the same amount of energy but often the provider doesn’t get paid for amps, it gets paid for watts – so providers are keen to see PF close to unity.

    2. In new luminaires with new capacitors (PF typically 0.8 lagging) the statement is quite true.  By adding a capacitive LED tube to the circuit the luminaire now has a capacitive (leading) PF and current increases because of this. However the LED tube is consuming about 60% less power (watts) than the fluorescent tube and control gear. Less power means less current so the additional current required from the leading PF is offset by the lower current required to drive the tube. At worst, the current may increase by 30% in a new luminaire with new and fully functioning capacitors. Read on to see if this is a problem.

    3. LED tube replacements are rarely performed on new luminaires.

    4. Capacitor performance degrades with time.  Old capacitors have normally either failed completely or are delivering little or no capacitance. After 10 years most capacitors will have degraded by up to 50% and a number will have failed.  Remembering that they are installed to offset the lagging power factor and so reduce current on the circuit, the fact that they are degrading and failing over time means that a 10 year old installation will now have increased current as the capacitance reduces.

    5. Lighting circuit designers know this.  The load on fluorescent circuits is designed to allow for “current creep” as the installation ages.  The lighting circuit is unlikely to reach capacity even with all capacitors failed.

    6. Back to paragraph b above (and thanks for reading on).  Does 30% increase in current matter? No – it is well within the original design parameters of the circuit which allowed for current creep over time.

    7. Noting paragraph c above, most LED tube upgrades are performed on older fluorescent luminaires in which capacitors are likely to be degraded and, because of less capacitance, the lagging PF has become worse and the current has increased.  Enter the new LED tube which is capacitive.  It is now enhancing or replacing the degraded or failed old capacitor taking power factor towards unity or perhaps into leading.

    8. The result is often a DROP in overall current on the circuit because it had already increased with age.


  4. If the current increases so does the power and so there is no energy saving

    1. If you are still with me after all this time, the statement is not true.  Current is not power and the increased current is reactive and wasted.  The power consumed by a 20 Watt LED tube is 20 watts.  In a 240 volt luminaire the current at unity PF is 0.0833amps. At PF 0.8 (leading or lagging) it is 0.11 amp and at PF 0.5 (leading or lagging) it is 0.166amps.  Regardless of the current it still only consumes 20 watts.


  5. If current on the lighting circuit increases the KVA demand charges from the retailer will increase.

    1. The statement is true if taken in isolation.

    2. Commercial buildings electrical loads are invariably inductive, caused by induction motors, power supplies and fluorescent and other gas discharge lighting.  So power factor, if uncorrected in a building, is normally lagging (inductive) and requires the retailer to provide more KVA than kW which is inefficient.  The retailer may charge demand charges to recover the cost of this inefficiency.

    3. Users correct inductive loads using Power Factor Correction Units (PFC).  These restore the overall PF in a building to unity so reducing KVA required to deliver KW required.

    4. A PFC unit consists of banks of capacitors with some control equipment allowing more capacitors to come on line as inductive load increases.

    5. Installing LED tubes in a building replaces an inductive load (fluorescent lights and ballasts) with a capacitive load.  The amount of capacitance entirely depends on the state of the PF capacitors in the old luminaires and will range from close to unity (all old capacitors have failed) to approximately 0.4 leading (all old capacitors are working perfectly). The latter example could be considered “worst case”. But is it? The ceiling has become capacitive and is now doing exactly what a PFC Unit does – offsetting other inductive loads and driving the overall inductive load of the site towards unity which is the ideal situation. A leading power factor in the ceiling may be very beneficial to the overall energy supply.


You have just read 1800 words supporting the techniques of installing LED tubes without any requirement to bypass ballasts or capacitors in old luminaire.  But there are some strong arguments as to why you SHOULD do so.  In the interests of a balanced discussion here they are:

  1. Upgrading to LED tubes is invariably undertaken to save power, money and abate greenhouse gas. If ballasts and capacitors are bypassed the luminaire cannot be returned to its original fluorescent configuration.  If they are not they can be easily changed back to fluorescent. If they do revert to fluorescent the energy, money and greenhouse gas savings will be lost.

  2. Because of this some Energy Saving Schemes have in the past made it compulsory to bypass ballasts.  Currently this is not the case but in the past it was a requirement because the schemes exist to abate greenhouse gas and reward participants based on 10 years of energy saving.  If the luminaire can readily be converted back to fluorescent the Scheme administrators saw this as a risk that 10 years of energy saving and abatement may not be achieved – so they didn’t allow it.

Cost/benefit is a primary consideration when businesses are considering investing capital. Armed with the information provided here decision makers hopefully are better able to assess the cost benefit of ballast and capacitor bypass when contemplating an upgrade with LED tubes.

Final word. There are two essential references that relate to best practice and safety which should be consulted in addition to the information above.  The first is the Energy Regulatory Authorities Council (ERAC) Guidelines of November 2011.  The second is the Australian Standard AS60598.2.1:2014 which supersedes the ERAC guidelines.  Neither have as a requirement to bypass ballasts and capacitors if installing LED tubes into fluorescent luminaires.



August 2015