Considering A Thermal Powerhouse As A 'System' And Building Inherent Reliability From The Gensets Up

Reliability is the probability (i.e. any number between 0 and 1) that a process/system/equipment/component/part will perform a specific function over a specific time period and under specific support conditions. Reliability is inherent in design! This means that reliability is a function of design and it is inbuild or designed into the process/system/equipment/component/part. The process/system/equipment/component/part represents the different physical asset hierarchical levels or structure. The designing in of reliability can happen at two levels in the hierarchical structure: from the equipment level and down; and from system level and up. From the equipment level and down is usually in the realm of OEMs (original equipment manufacturers) while from the system level and up is usually dictated by the end-users.

In PNG we are not a machine manufacturing country. So, from equipment level and down we often only liaise with the OEMs to improve machinery reliability once equipment is commissioned and repeated low characteristic life of components are noted in recurring failures through the equipment's operation. But as owners and end-users, we can have a direct impact on reliability at the process/system level during the upfront design stage of our process/system.

When it comes to improving the reliability of electricity supply, there are many options available but too often in PNG the tendency is to develop additional power generation and supply the existing power grid. For a vertically integrated power company like PPL, the option to reconfigure an existing powerhouse by adding gensets as standby capacity is also a viable option to consider but is often not pursued mostly because of upgrade costs (or simple lack of basic knowledge in the pragmatic ways of improving reliability). However, in situations where new power generation facilities will take longer to manifest and the immediate need is to stabilise power, building in redundancy or standby into the existing powerhouse may be the most viable and pragmatic thing to do in the short to medium term.

Improving reliability of an existing powerhouse may come down to building in redundancy or adding standby gensets and reconfigure the whole standard operating procedure to align with the new configuration. It starts by treating the powerhouse as the system with a 'mission function' (which could be stated as such for a powerhouse with x4 gensets: meeting the peak load of 4.7 MW annually for 15 years at an annual system availability of >80% and a load factor of >50%, it requires its x4 generators to be running as a system at >80% reliability annually).

As an example, let us consider the Boram Powerhouse in Wewak. Wewak is the largest diesel-only center within PNG Power Limited and in 2017 was reported to have had an average peak load of 4.2 MW. From the Post Courier report of 15th August 2024, the Boram Powerhouse currently has x4 gensets capable of generating a total of 4 MW (or each genset has 1 MW capacity).

To improve electricity supply reliability, we introduce a standby generator of 1 MW capacity to the existing x4 gensets to bring the total gensets up to x5 gensets each with a capacity of 1 MW. Our 'mission function' is still the same and that is to supply 4 MW at >80% availability over the next 15 years. This means that we have to configure the Boram Powerhouse to a 4/1 standby configuration where at any one time we will have x4 generators running and x1 is either on standby or undergoing some form of maintenance (either, planned maintenance or breakdown maintenance). The 'Parallel Non-Operating Redundancy Formula' already accounts for the conditions: hot/cold standby; planned maintenance; or breakdown maintenance.

So a 4 MW capacity thermal power station operating in a 4/1 standby configuration will have inherent reliability built into it. All the SOPs (standard operating procedures) and lifecycle costs of ownership including maintenance strategies and schedule power outages and etc for the power station can now be modified and optimised around this 4/1 standby configuration. This automatically guarantees electricity power reliability (i.e. assuming consistent and reliable diesel fuel supply).

Once the reliability of the powerhouse is improved by introducing the standby configuration, we can then proceed to optimise the operations and SOPs (standard operational procedures) of the powerhouse.

Now assuming that over 15 years from 2010, the power demand was predicted to grow by 2.2%. So Wewak, if it has 4.2 MW peak load in 2017 will now in 2025 have a peak load of 4.7 MW [i.e. 4.2 MW x (1 + 0.022 x 5 years)] and then in 15 years time in 2040 that peak load will be equal to about 6.251 MW [i.e. 4.7 MW x (1 + 0.022 x 15 years)]. Alternate energy options coming on line by the year 2040 could cater for the additional load increases. But from 2025 we decided to scale up our genset in the 4/1 standby configuration.

Assuming that the Boram Powerhouse in 2025 has x5 Hyundai HiMSEN engines installed and these are now scaled up to 2 MW each and still in this 4/1 standby configuration. That is, x5 gensets of 2 MW capacity each with SOPs reconfigured accordingly, that would give us an installed capacity of 8 MW (excluding the standby genset capacity of 2 MW) with a load factor of 58.75% (i.e. system average peak load of 4.7 MW divided by total installed capacity of 8 MW). This load factor is going to increase with system average peak load over time. The increase in load factor is because our installed capacity will be a constant which is now reliably guaranteed with the standby configuration and reconfigured SOPs and maintenance strategies. By having a 8 MW (i.e. 4 x 2 MW gensets - excluding standby of 2 MW) installed capacity, we are already covered in both reliability and effective utilisation for that anticipated increase over the next 15 year period (and if we were to scale up further with bigger capacity gensets in this 4/1 standby configuration, say 3 MW capacity gensets to account for system losses and reduce capacity due to wear and tear, that means Wewak will and can be easily covered for the next 30+ years).

Here is an example mathematics to demonstrate this reliability concept of introducing standby at the system level. Let us say that the Boram Powerhouse has its x5 generators. To fulfill its 'mission function' of meeting the peak load of 4.7 MW annually for 15 years at an annual system availability of >80% and a load factor of >50%, it requires its x4 generators to be running as a system at >80% reliability annually. The reliability of each gensets for the past 12 months are: G1 = 60%, G2 = 65%, G3 = 50%, G4 = 75% and G5 = 80%. How would you go about determining the total system reliability to feed into a reliability plan for the powerhouse and plan our maintenance strategies and lifecycle cost of ownership? Should we use direct averaging or 'block diagram modelling'? If we direct average (like most sites and people do) then chances are that we could be understating our true inherent powerhouse reliability (and availability) capacity in these proportions: Path1 = 62.5%, Path2 = 63.75%, Path3 = 70%, Path4 = 66.25% and Path5 = 67.5%. The correct system reliability in this 4/1 standby configuration is however 99.65%.

Here is the maths: 1 - (1 - 0.6) x (1 - 0.65) x (1- 0.5) x (1 - 0.75) x (1 - 0.8) = 99.65%

This is the mathematical proof that by introducing standby (in whatever the configuration this may be), we would have immensely improved reliability.