Scribd
Upload
3 views3 pages

Energy Project

The project focuses on optimizing real-time energy markets using data from buses, branches, and generators. It addresses the challenge of meeting demand under variable conditions, particularly wind variability, by modifying flow conservation constraints and introducing a penalty for lost demand. The project includes tasks such as solving the optimization problem, analyzing the impact of wind variability on costs, and investigating the potential for increasing generator capacity to reduce public costs.

Uploaded by

Nathalie He
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views3 pages

Energy Project

The project focuses on optimizing real-time energy markets using data from buses, branches, and generators. It addresses the challenge of meeting demand under variable conditions, particularly wind variability, by modifying flow conservation constraints and introducing a penalty for lost demand. The project includes tasks such as solving the optimization problem, analyzing the impact of wind variability on costs, and investigating the potential for increasing generator capacity to reduce public costs.

Uploaded by

Nathalie He
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Project: Real-time Energy Markets

1 Introduction
This project builds on the energy writeups we have covered in the course, with some simplifications.
The raw data is provided in the three csv files that come with this project (Buses, Branches, and
Generators). Note: all indexing is done starting from 0 rather than 1 (i.e., there is a bus 0, a branch
0 and a generator 0).
Data is as follows:

• Generators. For each generator h we provide its ID, the bus that it is housed at, the value
Uh and the cost function Ch (Phg ). This cost function is piecewise linear and given by three
parameters σ1,h , σ2,h and Xh , with the following interpretation:
Ch (Phg ) = σ1,h Phg , when 0 ≤ Phg ≤ Xh (1)
= σ1,h Xhg + σ2,h (Phg − Xhg ) when Xh ≤ Phg ≤ Uh (2)
Note: the cost function is convex, meaning σ1,h ≤ σ2,h (or we have a data error). Wind
generators have zero operating cost.
• Buses. For each bus k we have its demand Pkd .
• Branches. For each branch we provide its ID, its “from” and “to” buses, its y parameter,
and its limit.

In the attached data, the network has 1814 buses, 2023 branches and 395 generators.

1.1 First enhancement


As stated, the problem may be infeasible because the demand cannot be met (too much demand,
for example). In fact in this study we will consider scenarios where data changes (in particular
wind variability) causes the problem to become infeasible. To handle that, the flow conservation
constraint (2) of the energy document,
X X X g
fℓ − fℓ = Ph − Pkd , (3)
ℓ∈F (k) ℓ∈T (k) h∈G(k)

is modified as follows:
Phg − (Pkd − Sk ),
X X X
fℓ − fℓ = (4)
ℓ∈F (k) ℓ∈T (k) h∈G(k)

Here, Sk is a new variable; it models demand that is being ’lost’ or ’shed’. This quantity should be
kept as small as possible. We impose
0 ≤ Sk ≤ Pkd , (5)
and the objective function becomes accordingly, we add to the objective a penalty term:
P g 6
P
Minimize h∈G Ch (Ph ) + 10 k Sk (6)
For completeness, we also write all the constraints

1
1. For any line ℓ = (i, j)

fℓ = yℓ (θi − θj ) (7)

2. For any bus k,

Phg − (Pkd − Sk ),
X X X
fℓ − fℓ = (8)
ℓ∈F (k) ℓ∈T (k) h∈G(k)

0 ≤ Sk ≤ Pkd , (9)

3. For any generator h,

0 ≤ Phg ≤ Uh (10)

4. For any line ℓ,

|fℓ | ≤ Lℓ . (11)

2 The project
2.1 First task
Solve the real-time problem, i.e. the optimization problem given above. Is any demand lost?

2.2 Second task: pricing of demand


Suppose we solve the above problem. The way that power grid economics works (in a somewhat
simplified manner) is that the public, indirectly, pays
∗
P
cost = bus k πk (dk − Sk ) (12)

where the sum is over all the buses, and for each bus k,

πk∗ = Optimal dual variable for constraint (8). (13)

In this part of the project we study the impact on cost of wind variability. To that effect,
1. Assume that the ’U’ quantity for ’wind’ generators (indicated as such in ’generators.csv’) is
in fact the expectation of maximum available power.

2. Assume that the output of wind generators is multivariate Gaussian, with mean as just
indicated, and covariance as per the matrix ’covariance.txt’.

3. Repeat the following experiment at least 1000 times: sample output for the wind farms as
just indicated (please remember that output cannot be negative). Solve the resulting LP.
Then compute the cost as indicated above in formula (12).

4. What is the distribution of cost? Is it Gaussian? Discuss.

5. The public cost defined above becomes a random variable under the distribution I asked you
to simulate. Estimate the value-at-risk of this quantity at the 95 % level.

2
6. In the default case of the optimization problem introduce above (i.e. with wind output at the
original values), it should be the case that some (or many?) generators reach their output
limit. This suggests that at least some of these generators are more efficient than others –
hence if we could increase their capacity, the public cost would decrease. Investigate this
possibility: double the capacity of every non-wind generator whose output is at
the output limit as originally stated – does the VAR decrease? By how much?

Fri.Apr.18.205302.2025@freestyle

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy