extending a building-scale optimisation model to low

4th International Conference on Smart Energy Systems and 4th Generation District Heating 2018

#SES4DH2018

Extending a building-scale optimisation model to low-temperature district heating systems McKenna, R.*, Hagedorn, V.#, Mainzer, K.#

#Chair of Energy Economics, Karlsruhe Institute for Technology (KIT), Germany *DTU Engineering Management, Technical University of Denmark, Denmark

Literature overview and objective

4th International Conference on Smart Energy Systems and 4th Generation District Heating 2018 #SES4DH2018

Decentralized building supply • Akbari, Jolai & Ghaderi (2016) • Omu,

Choudhary & Boies (2013) • Wouters, Fraga & James (2015) • (Baetens et al. 2012) • Coninck et al. (2014) • Mehleri et al. (2013); (2012)

Centralized district supply • Yang, Zhang & Xiao (2015) • Orehounig, Evins & Dorer

(2015) • Walker et al. (2017) • Wu et al. (2018)

DH Network optimisation • Möller & Nielsen (2014) • Möller & Lund (2010) • Nielsen (2014) • Delangle et al. (2017) • Unternährer et al. (2017)

Study considers all three aspects

Objective: Comparison of centralized and decentralized energy supply systems in urban areas with different residential areas

General approach to model extension • Existing building-level MILP model extended to include heating grid • Directional graph • Building and heat generation plant shown as nodes • Forward and return flow shown in the model • Grid topology represented in the model by matrices


Dimensioning options for pipelines

• Each section is allocated a pipe diameter • Available pipe diameters are linked to various properties


Properties Data origin

Inner radius Exogenous

External radius Exogenous

Thermal conductivity of the insulation Exogenous

Max. heat flow Endogenous

Max. volume flow Endogenous

Max. & average flow velocity Exogenous

Material and installation costs Exogenous

Heat losses Endogenous

Pressure losses Endogenous

𝑈𝑈𝑅𝑅 =1

𝑟𝑟𝑅𝑅𝜆𝜆𝐷𝐷

ln 𝑟𝑟𝐴𝐴𝑟𝑟𝑅𝑅

+ 𝑟𝑟𝑅𝑅𝜆𝜆𝐵𝐵𝐵𝐵

ln� 4(ℎÜ + 𝑟𝑟𝐴𝐴)𝑟𝑟𝐴𝐴

+ 𝑟𝑟𝑅𝑅𝜆𝜆𝐵𝐵𝐵𝐵

ln 2(ℎÜ + 𝑟𝑟𝐴𝐴)𝑎𝑎 + 2𝑟𝑟𝐴𝐴

2+ 1

0,5

�

Calculation of heat losses 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑚𝑚𝑚𝑚𝑚𝑚,𝑅𝑅

𝐻𝐻𝐻𝐻𝑚𝑚𝐻𝐻,𝑙𝑙 = 𝑆𝑆𝑆𝑆𝑟𝑟𝑆𝑆𝑅𝑅,𝑙𝑙 ∗ ∆𝑇𝑇𝐿𝐿 ∗ 𝑈𝑈𝑅𝑅


𝑙𝑙𝑆𝑆𝑟𝑟𝑆𝑆𝑅𝑅 = 2 ∗ 𝜋𝜋 ∗ 𝑙𝑙𝑅𝑅 ∗ 𝑟𝑟𝑅𝑅

Relevant Surface

• Relative to 1m pipe length

Difference in temperature

• Flow & return constant

• Ground temperature constant

Heat transfer coefficient

• Installation distance & overlap height the same for all pipe diameters

𝛥𝛥𝑇𝑇𝐿𝐿 =𝑇𝑇𝑉𝑉𝐿𝐿 + 𝑇𝑇𝑅𝑅𝐿𝐿

2 − 𝑇𝑇𝐵𝐵𝐵𝐵

Balance Equation


Pipe diameters are defined on the basis of various restrictions

A

N-1

A-1

N N+1

T

T-1

Description of the model extension • Determination of the critical path in the model • Determine maximum case for critical path

– Maximum distance – Smallest pipe radii – Maximum case is subdivided for selection of head differences

• Consideration of the efficiencies in the model


Validation • Modifications of the model:

– Difference in altitutde – Network course – Load profils


Properties of the network

Location Böblingen, Germany

Length 2062 m

Connected houses 63

62

63

60

56

39

38

34

35

3637

33

32

31

3029

2826

27

23

25

24

22

19

20

54 53

4755

5748

49

45

46

50

51

52

44

41

40

43

42

17

16

13

15

14

128

9

10

7 65

4

3

2

1

21

11

58

59

31

30

2926

25

16

17

15

1312

10

14

11

9

2420

23

2221

32

18

19

8

7

6

5

33

3

2

1

34

2728

4

61

18

20m

Validation results


Comparison point Model Existing network

Annual heat losses 7,11% 25%

• Use of smaller pipe diameters in the model • Lower flow temperature in the model • Better insulation in the model

Validation results




Deviation of the installed boiler capacity Model 30 % lower

Deviation of the installed pump capacity Model 4% lower

• Differences in the heat demand of the networks (lower peak load in the model)

• Reduced heat losses mean that less heat generally has to be generated • Perfect foresight of the model

Validation results




Deviation of the installed boiler capacity Model 30 % lower

Deviation of the installed pump capacity Model 4% lower

Deviation of the annual heat production Model 21 % lower

Validation results: heat flow


0

50

100

150

200

250

300

350

400

00:0

0:00

01:0

0:00

02:0

0:00

03:0

0:00

04:0

0:00

05:0

0:00

06:0

0:00

07:0

0:00

08:0

0:00

09:0

0:00

10:0

0:00

11:0

0:00

12:0

0:00

13:0

0:00

14:0

0:00

15:0

0:00

16:0

0:00

17:0

0:00

18:0

0:00

19:0

0:00

20:0

0:00

21:0

0:00

22:0

0:00

23:0

0:00

kWh

Zeit [h]

Wärmefluss im Netz für Wintertag mit maximalem Bedarf (Validierungsnetz)

Wärmeverluste Mindestfluss QuartierswärmebedarfSpeicheranteil im Netz Gesamter Wärmefluss Wärmeproduktion des Kessels

Validation results: pump


0

50

100

150

200

250

300

0

0,5

1

1,5

2

2,5

3

3,500

:00

01:0

0

02:0

0

03:0

0

04:0

0

05:0

0

06:0

0

07:0

0

08:0

0

09:0

0

10:0

0

11:0

0

12:0

0

13:0

0

14:0

0

15:0

0

16:0

0

17:0

0

18:0

0

19:0

0

20:0

0

21:0

0

22:0

0

23:0

0

Qua

rtier

swär

meb

edar

f [kW

h]

Pum

plei

stun

g [k

W]

Zeit [h]

Pumpenbetrieb am maximalen Wintertag

Pumpleistung 2. Pumpe Pumpleistung 1. Pumpe Quartierswärmebedarf

Discussion • Question of over-specification… • Assuming not over-specified:

– Runtime reduction by e.g. decomposition methods – Increase the temporal scope of the model

- currently only 8 days – Consider more technologies – Better pump linearization trough SOS2-Constraints – Consider more realistic plant operating times

– Possibility to allocate individual capacities to the houses


Summary and Conclusions

• MILP model for DH network layout and operation, given demand sinks and network topology

• Possible to model and compare centralized and decentralized systems (not shown here)

• Validation shows deviations from empirical data are plausible and system operation is realistic

• Further work should improve the pump and plant operation, e.g. part load efficiencies and ramp rates


Thank you for your attention!


extending a building-scale optimisation model to low

Documents