Q10Power System Planning

(a) Operating and Maintenance Cost of a Candidate Plant

The operating and maintenance (O&M) cost of a candidate generating plant is one of the key economic parameters, alongside capital cost and fuel cost, used to characterize each candidate plant type in generation expansion planning studies. O&M cost represents the ongoing expenditure required to keep the plant running and properly maintained over its operating life, and is conventionally split into two components based on how the cost varies with the plant's operating level.

• Fixed O&M cost: costs that are incurred regardless of how much energy the plant actually generates, such as permanent staffing and administrative overheads, routine scheduled inspection and preventive maintenance, insurance, and land/property-related charges. Fixed O&M is usually expressed in currency per unit of installed capacity per year (e.g., cost per kW-year).
• Variable O&M cost: costs that vary directly with the amount of energy actually produced, such as consumables, lubricants, wear-and-tear-related component replacement, and maintenance activity that is proportional to operating hours or number of starts/stops. Variable O&M is usually expressed in currency per unit of energy generated (e.g., cost per MWh).

In least-cost expansion planning studies, both fixed and variable O&M costs of every candidate plant type are combined with its capital cost (annualized) and fuel cost (based on heat rate and fuel price) to compute the total annual cost associated with including that plant in a candidate expansion configuration. Accurately estimating O&M cost is important because candidate technologies differ significantly in this respect - for example, a gas turbine peaking plant typically has low capital cost but relatively higher variable O&M and fuel cost per unit of energy, making it economical only for low-utilization peaking duty, whereas a base-load coal or nuclear plant has higher capital cost but lower variable O&M and fuel cost per unit, making it economical for continuous high-utilization operation. The dynamic-programming based optimization in tools such as WASP therefore relies on accurate fixed and variable O&M cost data for every candidate plant type to correctly rank and select the least-cost combination of technologies for each year of the study horizon.

(b) Minimum Assured Reliability Constraints via Optimization/Programming

In generation expansion planning, cost minimization alone is not a sufficient objective, because a plan that is cheapest in pure cost terms might not provide an acceptable level of supply reliability to consumers. To address this, the least-cost optimization problem is formulated with an explicit minimum assured reliability constraint, ensuring that whichever expansion sequence is chosen as least-cost, it must also guarantee that the system's reliability index does not fall below (or, equivalently, the risk index does not exceed) a policy-specified minimum acceptable standard in every year of the study horizon.

• Reliability index selection: a suitable probabilistic reliability measure is chosen, most commonly loss of load probability (LOLP), expressed as the expected number of days or hours per year that available capacity may be insufficient to meet demand, or expected energy not served (EENS) in MWh per year.
• Threshold specification: a maximum acceptable value of the chosen reliability index (e.g., LOLP not exceeding one day in ten years, or an equivalent standard) is specified by the utility or regulator as a hard constraint on the optimization.
• Constraint embedding in dynamic programming: within the dynamic-programming based optimization (as used in packages such as WASP), the algorithm evaluates each candidate configuration's reliability performance (via probabilistic simulation, e.g., MERSIM in WASP) at every stage and rejects/penalizes any configuration whose reliability index fails to meet the required minimum standard, before comparing surviving configurations purely on cost.
• Trade-off exposed to the planner: the resulting formulation lets the planner see the additional cost incurred as a function of the reliability standard imposed, allowing informed policy decisions about the appropriate trade-off between system cost (and hence tariff impact) and the assured level of reliability for consumers.
• Outcome: the final expansion plan selected is the minimum-cost sequence among only those candidate sequences that satisfy the minimum assured reliability constraint in every year, ensuring that reliability is never compromised purely to reduce cost.

This constrained optimization approach - minimizing cost subject to a hard reliability floor - is the standard formulation used in practice, since it directly reflects the regulatory and social expectation that electricity supply must meet a guaranteed minimum quality of service, with cost minimization operating only within the space of plans that already satisfy this reliability requirement.

The block-diagram representation of the least-cost optimization problem highlights that the reliability simulation and cost-calculation stages are not independent: a configuration with lower installed capacity will generally show lower capital cost but higher expected-unserved-energy cost due to poorer reliability, and it is only by combining both effects into the single total-cost objective function that the dynamic-programming search can correctly identify the configuration that minimizes overall cost while still respecting the explicit reliability constraint applied at every stage of the recursion.

Both operating and maintenance cost estimation for candidate plants, and the imposition of minimum assured reliability constraints, ultimately serve the same underlying purpose within the least-cost planning framework: ensuring that the recommended expansion plan reflects the true, complete lifecycle cost of each technology option (not just its visible capital cost) and that the plan selected on this basis still delivers a guaranteed minimum quality of supply to consumers, rather than allowing an artificially low reported cost, achieved by understating O&M expense or by tolerating excessive unreliability, to distort the apparent ranking of competing expansion alternatives.

Operating and maintenance cost estimates for candidate plants are typically benchmarked against actual historical O&M expenditure of comparable existing units of the same technology and size, adjusted for inflation and any known technology improvements, since candidate-plant cost data supplied by equipment vendors at the pre-commitment stage can understate real-world O&M costs observed once a plant has been in service for several years.

Operating and maintenance cost projections for a candidate plant are also affected by the plant expected dispatch pattern within the optimized system: a plant selected for base-load duty accumulates operating hours (and hence variable O&M cost and component wear) much faster than an equivalent plant selected mainly for occasional peaking duty, so the least-cost optimization must jointly determine both which plants to add and how they will actually be dispatched, since the assumed dispatch pattern directly affects the total O&M cost that should be attributed to that plant over the study horizon.

The imposition of a minimum assured reliability constraint also has a direct and quantifiable cost, often referred to as the cost of reliability, which planners compute by re-solving the least-cost optimization at several different reliability threshold levels and comparing the resulting total system cost; presenting this cost-versus-reliability trade-off curve to regulators and policymakers is a standard and valuable output of the least-cost planning process, since it makes explicit the additional expenditure required for each incremental improvement in guaranteed supply reliability.