Printer Friendly

Distributionally robust self-scheduling optimization with C[O.sub.2] emissions constraints under uncertainty of prices.

1. Introduction

Generation self-scheduling in a pool-based electricity market has been recently studied in the power systems literature [13]. The self-schedules are required in developing successful bidding strategies and constructing hourly bidding curves for consideration by the independent system operator. In order to obtain successful generation bids, the generation companies have to self-schedule their unit by maximizing the expected profit based on the forecasted location marginal prices and accounting for the network security constraints. For a price-taker generation and a single time period, the generation schedule x is obtained from the following deterministic self-scheduling problem:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (1)

where x is an (m x 1) column vector containing the generation schedule, [xi] is an (m x 1) column vector of locational marginal prices (LMPs), X [subset] [R.sup.m] is the feasible region which is a convex and compact set, and [C.sub.i]([x.sub.i]) is the generation quadratic cost function:

[C.sub.i] ([x.sub.i]) = [a.sub.i] + [b.sub.i][x.sub.i] + [c.sub.i][x.sup.2.sub.i] i = 1, ..., m. (2)

The issue of interest for this work is that, since the electricity prices are of stochastic nature, the generation company cannot be certain about the revenue. Measuring the underlying risk due to this uncertainty is crucial not only for assessing profitability but also for generation self-scheduling. Stochastic programming can effectively describe self-scheduling problems in uncertain environments. Unfortunately, although the self-scheduling problem is a convex optimization problem, to solve it one must often resort to Monte Carlo approximations, which can be computationally challenging. A more challenging difficulty that arises in practice is the need to commit to a distribution given only limited information about the stochastic parameters [4].

In an effort to address these issues, robust formulations for self-scheduling problems were proposed; see [2, 3]. Jabr [2] considers a generation self-scheduling model based on a worst-case conditional robust profit with partial information on the probability distribution of prices. It is assumed that the nominal distribution and a set of possible distributions were given. Uncertainty of prices is represented by box and ellipsoidal uncertainty sets. However, in practice, true probability distribution of prices cannot be known exactly. Their solution can be misleading when there is ambiguity in the choice of a distribution for the random prices. Recently, Delage and Ye [4] proposed a distributionally robust optimization model that describes uncertainty in both the distribution form and moments (mean and covariance matrix). By deriving a new form of confidence region for the mean and the covariance matrix of a random vector, it was showed how the proposed distribution set can be well justified when addressing data-driven problems (i.e., problems where the knowledge of the random parameters is solely derived from historical data).

On the other hand, as a major energy-saving industry, power industry has implemented energy-saving generation dispatching. Apart from security and economy, low carbon will be the most important target in power dispatch mechanisms. It becomes a common target for the global power industry to build a more safe, economic, and low-carbon power system [5]. So more attention should be paid to the reduction of C[O.sub.2] emission in power system operation [5, 6].

The main contribution of this paper is twofold. First, by considering a power system with many thermal power generators which use different petrochemical fuels (such as coal, petroleum, and natural gas) to produce electricity, respectively, we establish a self-scheduling model based on the forecasted locational marginal prices, particularly taking into account C[O.sub.2] emission constraint, C[O.sub.2] emission cost, and unit heat value of fuels. This problem is important and timely as world leaders and international organizations discuss the roles and responsibilities of each country and sector of economic activity in the path towards a sustainable future. Second, motivated by the work of Delage and Ye [4], we propose a distributionally robust self-scheduling optimization model under uncertainty in both the distribution form and moments of the locational marginal prices, where the knowledge of the prices is solely derived from historical data. Then we prove that the proposed robust self-scheduling model can be solved to any precision in polynomial time. These arguments are confirmed in a practical example on the IEEE 30 bus test system.

2. Robust Self-Scheduling Problem with Moment Uncertainty and C[O.sub.2] Emission Constraint

2.1. Power Dispatch with C[O.sub.2] Emission Constraint. In this section, we describe the power scheduling problem with C[O.sub.2] emission constraint.

Given is a thermal power system with M thermal power generators i [member of] M = {1, ..., m}, which use petrochemical fuels as their fuels, such as coal, petroleum, and natural gas. The quantity of C[O.sub.2] emission of each power generator i can be represented as

[E.sub.i] = [F.sub.i][h.sub.i], (3)

where [F.sub.i] is the quantity of the fuel which is consumed to produce electricity and [h.sub.i] ([greater than or equal to] 0) is the amount of C[O.sub.2] emissions by per unit of fuel complete combustion. The generating capacity of this power generator is

[x.sub.i] = [F.sub.i][p.sub.i][[phi].sub.i], (4)

where [p.sub.i] (> 0) is the unit heat value of the fuel and [[phi].sub.i] (> 0) is the energy conversion efficiency.

By (3) and 4), we get the carbon characteristic function of power generator i:

[E.sub.i] = [[h.sub.i]/[p.sub.i]] [1/[[phi].sub.i]] [x.sub.i]. (5)

The objective is the maximization of the expected profit based on the forecasted locational marginal prices [xi], particularly taking into account C[O.sub.2] emission constraints and C[O.sub.2] emission cost. The cost consists of the variable cost of electricity production of the thermal generators and the cost of C[O.sub.2] emission. So the power scheduling problem with C[O.sub.2] emission constraints can be modeled as the following deterministic self-scheduling problem:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (6)

where x is an (m x 1) column vector containing the generation schedule, [xi] is an (m x 1) column vector of forecasted locational marginal prices (LMPs), X is the feasible region, [delta] is the C[O.sub.2] emission price with the unit [pounds sterling]/ton, [bar.E] is the maximum allowable C[O.sub.2] emissions, and [C.sub.i]([x.sub.i]) is the generator quadratic cost function:

[C.sub.i] ([x.sub.i]) = [a.sub.i] + [b.sub.i][x.sub.i] + [c.sub.i][x.sup.2.sub.i], i = 1, ..., m, (7)

where [c.sub.i] > 0 for i = 1, ..., m.

Letting [D.sub.i]([x.sub.i]) [equivalent to] [delta]([h.sub.i]/[p.sub.i])(1/[[phi].sub.i]) [x.sub.i], by 5), problem (6) can be rewritten as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (8)

2.2. Robust Self-Scheduling Problem with Moment Uncertainty and C[O.sub.2] Emission Constraints. In practice, It is often the case that one has limited information about the locational marginal prices [xi] driving the uncertain parameters that are involved in the decision making process. The data samples may be not available, or the data samples may be unstable. In such situations, it might instead be safer to rely on estimates of the mean [[mu].sub.0] and covariance matrix [[summation].sub.0] of the random vector [xi], for example, using empirical estimates. However, we believe that, in such problems, it is also rarely the case that one is entirely confident in these estimates. For this reason, motivated by Delage and Ye [4], we propose representing the uncertainty using two constraints parameterized by [[gamma].sub.1] [greater than or equal to] 0 and [[gamma].sub.2] [greater than or equal to] 1:

[(E [[xi]] - [[mu].sub.0]).sup.T] [[summation].sup.-1.sub.0] (E [[xi]] - [[mu].sub.0]) [less than or equal to] [[gamma].sub.1], (9)

E [([xi] - [[mu].sub.0])[([xi] - [[mu].sub.0]).sup.T]] [??] [[gamma].sub.2][[summation].sub.0], (10)

where constraint (9) assumes that the mean of price [chi] lies in an ellipsoid of size [[gamma].sub.1] centered at the estimate [[mu].sub.0] and constraint (10) forces the centered second-moment matrix of [xi] to lie in a positive semidefinite cone defined with a matrix inequality. In other words, it describes how likely [xi] is to be close to [[mu].sub.0] in terms of the correlations expressed in Z0. Finally, the parameters [[gamma].sub.1] and [[gamma.sub.]2] provide natural means of quantifying one's confidence in [[mu].sub.0] and [[summation].sub.0], respectively.

Denote the distributional set as

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (11)

where U is the set of all probability measures on the measurable space ([R.sup.m], B), with B being the Borel [sigma]-algebra on [R.sup.m], and F [member of] [R.sup.m] is any closed convex set known to contain the support of F. The set [D.sup.1] (F, [[mu].sub.0], [[summation].sub.0], [[gamma].sub.1], [[gamma].sub.2]), which will also be referred to in shorthand notation as [D.sub.1], can be seen as a generalization of many previously proposed sets.

Next we will study worst-case expected results over the choice of a distribution in the distributional set [D.sub.1]. This leads to solving the distributionally robust self-scheduling optimization with moment uncertainty of prices (DRSSO):

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (12)

where [E.sub.F] [*] is the expectation taken with respect to the random vector [xi] given that it follows the probability distribution F [member of] [D.sub.1]. It is easy to see that DRSSO problem (12) is equivalent to the following problem:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (13)

We start by considering the question of solving the inner maximization problem of the problem (13) that uses the set [D.sup.1].

Definition 1. Given any fixed x [member of] X, let [PHI](x; [[gamma].sub.1], [[gamma].sub.2]) be the optimal value of the moment problem:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (14)

Applying the duality theory and robust optimization methods [7-10], by [4, Lemma 1], we can circumvent the difficulty of finding the optimal value of the problem 14).

Lemma 2. For a fixed x [member of] [R.sup.n], suppose that [[gamma].sub.1] [greater than or equal to] 0, [[gamma].sub.2] [greater than or equal to] 1, [[summation].sub.0] [??] 0. Then [PHI](v; [[gamma].sub.1], [[gamma].sub.2]) must be equal to the optimal value of the problem (15);

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (15)

where (A x B) refers to the Frobenius inner product between matrices, Q [member of] [R.sup.mxm] is a symmetric matrix, the vector q [member of] [R.sup.m], and r, t [member of] R. In addition, if [PHI](x; [[gamma].sub.1], [[gamma].sub.2]) is finite, then the set of optimal solutions to problem (15) must be nonempty.

Proof. Let h(v, [xi]) = -[[xi].sub.T]x + [[summation].sup.m.sub.i=1] [C.sub.i]([x.sub.i]) + [[summation].sub.m.sub.i=1] [D.sub.i] ([v.sub.i]). Then h(v, [xi]) is F-integrable for all F [member of] [D.sub.1] and the conclusions are followed by [4, Lemma 1]. This completes the proof.

The following result shows that the DRSSO problem (13) is a tractable problem.

Theorem 3. Assume that the set X is convex and compact. Then DRSSO problem (13) is equivalent to the following convex optimization problem;

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (16)

In addition, the DRSSO problem (13) can be solved to any precision [epsilon] in time polynomial in log(1/[epsilon]) and the size of x and [xi].

Proof. By Lemma 2, the DRSO problem 13) must be equal to the following problem:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (17)

Let

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (18)

Then [eta] [greater than or equal to] [[summation].sup.m.sub.i=1] [C.sub.i] ([x.sub.i]) + [[summation].sup.m.sub.i=1] [D.sub.i]([x.sub.i]), and the inequality (17) can be replaced by

[[xi].sup.T] Q[xi] + [[xi].sup.T]q + [[??].sup.T]x + r - [eta] [greater than or equal to] 0, [for all][xi] [member of] F. (19)

Note that the constraint in (19) can be written as the following LMI:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (20)

So DRSSO problem (13) is equivalent to the convex optimization problem (16).

Since f(x, [xi]) = -[[xi].sup.T]x + [[summation].sup.m.sub.i=1] [C.sub.i]([x.sub.i]) + [[summation].sup.m.sub.i=1] [D.sub.i]([x.sub.i] ) is convex in x and concave in [xi] and X is convex and compact, the assumptions in [4] are satisfied. So a straightforward application of [4, Proposition 2] shows that DRSSO problem (13) can be solved in polynomial time. This completes the proof.

If X is a convex polyhedron, then the convex optimization problem (16) is a semidefinite program that can be solved by SeDuMi conic optimizer.

3. Numerical Example

We present our simulation results on the IEEE 30 bus test system and get the results by using the SeDuMi conic optimizer running on an Intel Core i3-2350 M (2.30 GHz) PC with 2 GB RAM. There are 6 power generations with coal as their fuel in this system. Reference [11] gives the network and load data for this system. The generator data is listed in Table 1. A historical data set of 100 samples of prices vector [xi] is shown in [2, Table 2]. We assume that the generating units in Table 1 belong to the same generation company.

In implementing our method, the distributional set is formulated as [D.sub.1]([R.sup.6], [[mu].sub.0], [[summation].sub.0], 0.2, 2.3), where [[mu].sub.0] and [[summation].sub.0] are the empirical estimates of the mean and covariance matrix of prices vector [xi] shown in [2, Table 2].

We choose the parameters [[gamma].sub.1] and [[gamma].sub.2] based on some simple statistical analysis of a lot of experiments. We convert the calorific value of different kinds of coals to standard unit of coal and let the unit heat value of fuel p equal 8.13 (kW x h)/kg standard unit of coal, the amount of C[O.sub.2] emissions by per unit of fuel complete combustion h equal 2.62 kg C[O.sub.2]/kg standard unit of coal in uniform, and the maximum allowable C[O.sub.2] emissions E equal 150 ton.

If we fix [[gamma].sub.2] = 1.2 and let [[gamma].sub.1] vary from 0 to 2, it can be shown that the profit decreases when [[gamma].sub.1] increases from 0 to 1.2, and the profit is almost invariable after [[gamma].sub.1] > 1.2. Table 2 illustrates the generation self-scheduling result for [[gamma].sub.1] from 0 to 1 and [[gamma].sub.2] = 1.2.

The value of [[gamma].sub.1] reflects the stability of the electricity price. The electricity price is more instable when the value of [[gamma].sub.1] is higher. If generations produce too much electricity, they would face more risks. The decision makers would produce less electricity when the value of [[gamma].sub.1] is high. And due to the basic power needs, the generating capacity would tend to be stable if the value of [[gamma].sub.1] is too high. Figure 1 shows the result intuitively.

However, if we fix [[gamma].sub.1] and let [[gamma].sub.2] vary from 1 to 10, numerical results show that the profit is almost invariable. This implies that [[gamma].sub.2] is not sensitive to the model.

Now we study three kinds of fuels with different unit heat values. The unit heat value p of natural gas, coal, and oil is shown in [12, Table 1]. And we can get h, the quantity of C[O.sub.2] released by the unit fuel burnt fully, by some simple calculations. The relevant data is listed in Table 3. From this table, we can know that the unit heat value of natural gas is higher than that of coking coal and the quantity of C[O.sub.2] released by the unit natural gas is lower than that of it. Let [delta] = 0.2 [pounds sterling]/ton, [bar.E] = 135.0 ton. We assume the 6 buses all use one fuel as their power fuel; the three results are shown in Table 4.

The results show that using natural gas gets more economic benefits. This is because the unit natural gas can produce more quantity of heat and release fewer C[O.sub.2]. And the result of using oil is between natural gas and coking coal. In addition, we can find that the quantity of C[O.sub.2] emission of coking coal and the quantity of C[O.sub.2] emission of oil have reached the maximum allowable value, respectively.

For different power generators, we let their efficiency increase in turns. Numerical results show that the profits increase as their efficiency increases.

We consider the carbon price's effect in the following. Figure 2 illustrates the effect of carbon price [delta] on the generation self-scheduling. The upper part of the figure shows that the profit and C[O.sub.2] emission decrease when the carbon price [delta] increases with the limit of [bar.E]. And the lower part of the figure shows the result without the limit of [bar.E]. When [delta] is very small, the C[O.sub.2] emission is so large because there is no limit of [bar.E]. That is, [bar.E] makes a difference on reducing carbon emissions. When [delta] increases, the profit decreases. So government should consider the effect of carbon price when they set the carbon price.

We let the carbon price [delta] equal 0.2 [pounds sterling]/ton and 0.6 [pounds sterling]/ton, respectively, and [bar.E] range from 120 ton to 180 ton. Figure 3 illustrates this result. Due to the emission restriction decreases, the object f and C[O.sub.2] emission go up. And they tend to be stable along with the value of [bar.E] becoming larger. By the comparison between the upper and the lower part of Figure 3, the quality of C[O.sub.2] emission tends to reach an invariable value earlier as [delta] = 0.6 [pounds sterling]/ton. This shows that a suitable carbon tax is a good means to reduce carbon dioxide emissions.

Figure 4 demonstrates the change of generation self-schedule when [bar.E] increases. At first when [bar.E] increases, some generation schedule x increases. But when the [bar.E] achieves a point, the effect of restriction disappears and the generation schedules are invariable.

4. Conclusion

This paper studies worst-case profit self-schedules of price-taker generators with the constraints of C[O.sub.2] emission in pool-based electricity markets. A distributionally robust self-scheduling optimization model describes uncertainty of prices in both distribution form and moments (mean and covariance matrix), where the knowledge of the prices is solely derived from historical data. It is proved that the proposed robust self-scheduling model can be solved to any precision in polynomial time. These arguments are confirmed in a practical example on the IEEE 30 bus test system. Numerical results show that parameter [[gamma].sub.1] of mean is sensitive to the solution and parameter [[gamma].sub.2] of covariance is not sensitive to the solution. Through the comparison between the different fuels and energy conversion efficiencies, we find that the power stations should choose the fuels with lower h and higher p as their power fuels. And the energy conversion efficiency is also very important to the power stations; we should try to improve it. Through the comparison between the carbon prices [delta] and [bar.E], it is showed that the maximum allowable C[O.sub.2] emissions makes a difference in reducing carbon emissions and a reasonable carbon tax is a good means to reduce carbon dioxide emissions.

http://dx.doi.org/10.1155/2014/356527

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

This work was supported by the Hunan Provincial Natural Science Foundation of China (no. 14JJ2053).

References

[1] H. Yamin, S. Al-Agtash, and M. Shahidehpour, "Security-constrained optimal generation scheduling for GENCOs," IEEE Transactions on Power Systems, vol. 19, no. 3, pp. 1365-1372, 2004.

[2] R. A. Jabr, "Generation self-scheduling with partial information on the probability distribution of prices," IET Generation, Transmission & Distribution, vol. 4, no. 2, pp. 138-149, 2010.

[3] R. A. Jabr, "Robust self-scheduling under price uncertainty using conditional value-at-risk," IEEE Transactions on Power Systems, vol. 20, no. 4, pp. 1852-1858, 2005.

[4] E. Delage and Y. Ye, "Distributionally robust optimization under moment uncertainty with application to data-driven problems," Operations Research, vol. 58, no. 3, pp. 595-612, 2010.

[5] Q. Chen, C. Kang, Q. Xia, and D. S. Kirschen, "Mechanism and modelling approach to low-carbon power dispatch," Automation of Electric Power Systems, vol. 34, no. 12, pp. 18-23, 2010.

[6] C. Li, Y. Liu, Y. Cao, Y. Tan, C. Xue, and S. Tang, "Consistency evaluation of low-carbon generation dispatching and energy-saving generation dispatching," Proceedings of the Chinese Society of Electrical Engineering, vol. 31, no. 31, pp. 94-101, 2011.

[7] A. Ben-Tal, L. El Ghaoui, and A. Nemirovski, Robust Optimization, Princeton University Press, 2009.

[8] M. Sim, Robust Optimization, Operations Research, 2004.

[9] F. J. Fabozzi, P N. Kolm, D. A. Pachamanova, and S. M. Focardi, Robust Portfolio Optimization and Management, John Wiley & Sons, 2007

[10] D. Bertsimas, D. B. Brown, and C. Caramanis, "Theory and applications of robust optimization," SIAM Review, vol. 53, no. 3, pp. 464-501, 2011.

[11] O. Alsac and B. Stott, "Optimal load flow with steady-state security," IEEE Transactions on Power Apparatus and Systems, vol. 93, no. 3, pp. 745-751, 1974.

[12] T. Hongqing, "Research on energy transformation rate in coal chemical plant," Chemical Engineering Design Communications, vol. 37, no. 4, pp. 5-13, 2011.

Minru Bai (1) and Zhupei Yang (2)

(1) College of Mathematics and Econometric, Hunan University, Changsha, Hunan 410082, China

(2) China CITIC Bank, Changsha Shuyuan Road Branch, Changsha, Hunan 410015, China

Correspondence should be addressed to Minru Bai; minru-bai@163.com

Received 7 April 2014; Accepted 18 May 2014; Published 2 June 2014

Academic Editor: Nan-Jing Huang

TABLE 1: Generator data.

Bus      [x.sub.min],    [x.sub.max],   Cost coefficients
number        MW              MW

                                           a, [pounds
                                           sterling]/h

1             50             200                0
2             20              80                0
5             15              50                0
8             10              35                0
11            10              30                0
13            12              40                0

Bus      Cost coefficients
number

          b, [pounds     a, [pounds
         sterling]/MWh   sterling]/h
                         M[W.sup.2]h

1            2.00          0.00375
2            1.75          0.01750
5            1.00          0.06250
8            3.25          0.00834
11           3.00          0.02500
13           3.00          0.02500

TABLE 2: Generation self-scheduling result
for [[gamma].sub.1] from 0 to 1 ([[gamma].sub.1] = 1.2).

[[gamma].sub.1.]   Bus 1, MW   Bus 2, MW   Bus 5, MW   Bus 8, MW

0                   135.00       35.37       16.45       35.00
0.2                 103.36       33.44       16.20       35.00
0.4                  91.42       32.45       16.08       35.00
0.6                  82.87       31.62       15.98       35.00
0.8                  76.13       30.88       15.90       35.00
1                    70.55       30.20       15.82       35.00

[[gamma].sub.1.]   Bus 11, MW   Bus 13, MW    Profit,
                                              [pounds
                                             sterling]

0                    30.00        38.67       229.52
0.2                  30.00        36.77       188.78
0.4                  30.00        35.94       174.14
0.6                  30.00        35.29       163.71
0.8                  30.00        34.73       155.41
1                    30.00        34.23       148.44

TABLE 3: Relevant data of fuels.

Power fuel       [h.sub.i],       [p.sub.i], Kwh/kg
              kg(C[O.sub.2])/kg

Coking coal         3.04                7.90
Natural gas         2.18                10.81
Oil                 3.06                11.61

TABLE 4: The result of using one kind of fuel
(S = 0.2 [euro]/ton, [bar.E] = 135.0 ton).

Bus number    1, MW   2, MW   5, MW   8, MW   11, MW

Natural gas   99.24   31.18   15.47   35.00   30.00
Oil           68.42   25.36   15.00   35.00   30.00
Coking coal   50.00   20.00   15.00   24.88   15.55

Bus number    13, MW   Profit, [eurp]   C[O.sub.2], t

Natural gas   35.31        174.12           119.1
Oil           31.10        154.89           135.0
Coking coal   14.90        113.84           135.0
COPYRIGHT 2014 Hindawi Limited
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Research Article
Author:Bai, Minru; Yang, Zhupei
Publication:Journal of Applied Mathematics
Article Type:Report
Date:Jan 1, 2014
Words:4163
Previous Article:Detecting runoff variation of the mainstream in Weihe River.
Next Article:Predicting future deterioration of hydraulic steel structures with Markov chain and multivariate samples of statistical distributions.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters