MLF Class

Description: marginal loss factor equation
Detail:

See also MLF Property Reference for a detailed list of properties for this class of object.

This class is specifically designed for the Australian NEM. NEM loss equations include terms for regional demand as well as flow. The simulator (like the NEM market-clearing engine) approximates the loss function with a multi-tranche linear approximation.

You enter the MLF equations directly into the model using special MLF objects, and the simulator will calculate the loss function accounting for all factors including region demand, and from those function generates a linear approximation. This approach has the advantage that loss functions on the interregional interconnectors are calculated every period, ensuring that the simulated losses and interregional prices are as close as possible to an emulation of the market. There is no significant additional overhead or iteration required, so this is a also very efficient approach.

Example:

Class Name Category Description
MLF LF Murray 330 MLF Loss factor equation (Murray 330 referred to Thomastown 66)
MLF LF South Pine 275 MLF Loss factor equation (South Pine 275 referred to Sydney West 330)
MLF LF Sydney West 330 MLF Loss factor equation (Sydney West 330 referred to Murray 330)
MLF LF Torrens Island 66 MLF Loss factor equation (Torrens Island 66 referred to Thomastown 66)

Parent Class Child Class Collection Constraint Child Name
MLF Region Regions LF Murray 330 NSW1
MLF Region Regions LF Murray 330 SA1
MLF Region Regions LF Murray 330 VIC1
MLF Region Regions LF South Pine 275 NSW1
MLF Region Regions LF South Pine 275 QLD1
MLF Region Regions LF Sydney West 330 NSW1
MLF Region Regions LF Sydney West 330 VIC1
MLF Region Regions LF Torrens Island 66 SA1
MLF Region Regions LF Torrens Island 66 VIC1
MLF Node Reference Node LF Murray 330 Murray 330kV
MLF Node Reference Node LF South Pine 275 South Pine 275kV
MLF Node Reference Node LF Sydney West 330 Sydney West 330kV
MLF Node Reference Node LF Torrens Island 66 Torrens Island 66kV
MLF Line Line LF Murray 330 V-SN
MLF Line Line LF South Pine 275 NSW1-QLD1
MLF Line Line LF Sydney West 330 SNOWY1
MLF Line Line LF Torrens Island 66 V-SA

Parent Class Child Class Collection Parent Name Child Name Property Value
System MLF MLFs NEM LF Murray 330 Intercept 1.0163
System MLF MLFs NEM LF South Pine 275 Intercept 1.0027
System MLF MLFs NEM LF Sydney West 330 Intercept 1.0144
System MLF MLFs NEM LF Torrens Island 66 Intercept 0.9849
MLF Line Lines LF Murray 330 V-SN Flow Coefficient 0.000095342
MLF Line Lines LF South Pine 275 NSW1-QLD1 Flow Coefficient 0.00023474
MLF Line Lines LF Sydney West 330 SNOWY1 Flow Coefficient 0.000080175
MLF Line Lines LF Torrens Island 66 V-SA Flow Coefficient 0.00038885
MLF Region Regions LF Murray 330 NSW1 Demand Coefficient -0.0000028101
MLF Region Regions LF Murray 330 SA1 Demand Coefficient -0.0000012838
MLF Region Regions LF Murray 330 VIC1 Demand Coefficient 0.000001691
MLF Region Regions LF South Pine 275 NSW1 Demand Coefficient 0.0000017283
MLF Region Regions LF South Pine 275 QLD1 Demand Coefficient 0.0000010083
MLF Region Regions LF Sydney West 330 NSW1 Demand Coefficient -0.0000059551
MLF Region Regions LF Sydney West 330 VIC1 Demand Coefficient -0.0000011774
MLF Region Regions LF Torrens Island 66 SA1 Demand Coefficient -0.0000099266
MLF Region Regions LF Torrens Island 66 VIC1 Demand Coefficient 0.0000022497

Interconnector losses are attributed to the sending and receiving regions in certain proportions. In the Australian NEM, these proportions are fixed constants. Line objects include the property Loss Allocation. By default, this allocation is 0.5, which means losses are attributed evenly to the sending and receiving nodes. The value may be set to any value between zero and one with values above 0.5 meaning that more losses are assigned to the receiving node.

Example:

Line Property Value
N-Q-MNSP1 Loss Allocation 0.23
NSW1-QLD1 Loss Allocation 0.5
SNI Loss Allocation 0.5
SNOWY1 Loss Allocation .95
V-SA Loss Allocation 0.6
V-S-MNSP1 Loss Allocation 0.6
V-SN Loss Allocation 1

Note that losses for the SNOWY1 interconnector are 'normally' allocated in part to Snowy (10%) and NSW (90%), and likewise for V-SN, but for medium term modelling purposes, where no load is attributed to the Snowy region, these allocation factors should be consistent in not allocating any load to Snowy. For short-term modelling, where load are input for Snowy the 'normal' factors should be used (Snowy1 0.95/0.05, V-SN 0.10/0.90).

The NEM uses 10-piece linear loss functions across the entire range of feasible flows, notionally treating maximum import as the 'zero-point' for the approximation. The Transmission setting Max Loss Tranches determines how many loss tranches are used in each direction of flow. Thus, setting this value to 5 will replicate the approach used in the NEM.

The following pictures capture the dynamic MLF and loss approximation over a one-year simulation. There are eight graphs, two for each interconnector. The first chart in the set plots line flow versus loss, treating flows in the 'back' direction as notionally negative. The second chart shows the marginal loss factor (as reported under Marginal Loss Factor). This graph shows clearly the steps in the linear approximation. The green points are the MLF calculated using the 'raw' MLF equations above. The spread of points on each chart shows the amount of variation in the MLF equations due to the dynamic demand elements.

Figure 1: NSW-QLD Dynamic Losses (1-year period)

Figure 2: NSW-QLD MLF (10-step Linear Approximation versus Calculated MLF)

Figure 3: Snowy Dynamic Losses (1-year period)

Figure 4: Snowy MLF (10-step Linear Approximation versus Calculated MLF)

Figure 5: V-SN Dynamic Losses (1-year period)

Figure 6: V-SN MLF (10-step Linear Approximation versus Calculated MLF)

Figure 7: V-SA Dynamic Losses (1-year period)

Figure 8: V-SA MLF (10-step Linear Approximation versus Calculated MLF)