Brief Summary
Okay ji, so this video explains how to find the most economical way to run power generating units using dynamic programming. It also talks about how important it is to keep the power supply reliable and how that affects decisions about which units to use. Key points include:
- Using a recursive relation to find the minimum operating cost.
- Balancing the cost of running different units with the need for a reliable power supply.
- Understanding concepts like spinning reserve and maximum tolerable insecurity level (MTIL).
Introduction to Dynamic Programming in Unit Commitment
So, basically, we're looking at how to figure out the cheapest way to generate a certain amount of power (x megawatts) using N number of generating units. We define FNx as the minimum cost in rupees per hour to generate x megawatts using N units. And fNy is the cost to generate y megawatts using the Nth unit. The main idea is to find the right combination of units to minimize the overall cost.
Recursive Relation and Computational Complexity
The application of dynamic programming gives us a recursive relation: FNx = minimize y [fNy + FN-1(x-y)]. This equation helps us find the best combination of units to use. While the equation itself looks simple, solving it can be computationally intensive, especially when you have to consider many different combinations of units and load levels.
Sample System and Cost Curve Parameters
Let's consider a system with four generating units. Each unit has a minimum generating capacity of 1 megawatt and a maximum of 12 megawatts. The cost of generation is given by a quadratic equation: aPG^2 + bPG, where PG is the power generated. The values of 'a' and 'b' are different for each unit. The goal is to determine which units should be used to meet a load of 9 megawatts in the most economical way.
Finding the Optimal Combination for Two Units
First, we consider only one unit (N=1). Then F1(x) = f1(x) = a1PG1^2 + b1PG1. If unit 1 generates all 9 megawatts, the cost can be calculated using the given values of a1 and b1. Next, we consider two units (N=2). We need to find the minimum cost by considering all possible combinations of power generated by the two units, such that their total output is 9 megawatts. After computing all combinations, we find that the minimum cost is achieved when unit 2 generates 2 megawatts and unit 1 generates 7 megawatts.
Adding the Third and Fourth Units
When we add the third unit, we need to compute F3(9) by considering all combinations of power generated by the three units. This involves finding the minimum cost for F2(8), F2(7), and so on, which requires considering all possible combinations for units 1 and 2. After all calculations, it turns out that the minimum cost is still achieved when only units 1 and 2 are operating, with unit 2 generating 2 megawatts and unit 1 generating 7 megawatts. Similarly, adding the fourth unit doesn't change the optimal combination.
Reliability Considerations in Unit Commitment
Now, let's talk about reliability. While using a single large generating unit (e.g., 500 MW) might be cheaper, it's less reliable. If that unit goes out, you lose a lot of power. It's better to have multiple smaller units (e.g., five 100 MW units). This way, if one unit fails, you still have the others to supply power. Electric utilities need to balance economy and reliability to ensure a continuous and quality power supply.
Spinning Reserve and System Security
To handle unexpected failures, power systems use something called "spinning reserve." This is extra generating capacity that's already online and ready to quickly increase output if needed. The spinning reserve provides a margin over the load requirements. The probability of a unit outage increases with operating time. Also, loads are never known with complete certainty.
Modeling Unit Performance and Reliability Parameters
To analyze reliability, we can model unit performance using parameters like "mean time to failure" (MTTF) and "mean time to repair" (MTTR). MTTF is the average time a unit operates before failing, and MTTR is the average time it takes to repair a failed unit. From these, we can calculate failure rate (lambda) and repair rate (mu). These rates are estimated from past data.
Patton's Security Function and Security Constraint Optimal Unit Commitment
Patton's security function (S) estimates the probability that the available generation capacity is less than the system load. S = Σ(pI * rI), where pI is the probability of the system being in state I, and rI is the probability that state I causes a breach of system security. We want to keep S below a certain maximum tolerable insecurity level (MTIL). If S exceeds MTIL, we need to modify the unit commitment schedule by bringing in the next most economical unit.
Example of Security-Constrained Unit Commitment
Consider a system with a daily load curve and an economically optimal unit commitment table. We need to check if the unit commitment is secure in every period of the load curve. For example, during a period with a minimum load of 5 megawatts, the unit commitment table might say to operate only unit 1. However, if we calculate the security function S and find that it exceeds the MTIL, we need to bring in the next most economical unit (unit 2) to improve the system's security.
