Methodology

Research articles · Dec 31, 2016

Constructing a synthetic city for estimating spatially disaggregated heat demand

M. Esteban Muñoz H., Ivan Dochev, Hannes Seller, Irene Peters

Technical Urban Infrastructure Systems Group, HafenCity University, Germany

Volume 9(3) Winter 2016

Cite as: M. Muñoz H., I. Dochev, H. Seller, I. Peters; 2016; Constructing a synthetic city for estimating spatially disaggregated heat demand; International Journal of Microsimulation; 9(3); 66–88. doi: 10.34196/ijm.00143

Abstract

We present a procedure for creating a spatially referenced building stock with population living therein “a synthetic city” for the case of Germany. The level of spatial disaggregation is the European NUTS–3 level for which data from the newest census (2011) exist. Our application is on the estimation of heat demand.

We use the German microcensus (2010) which contains both: (a) detailed sociodemographic characteristics of individuals and (b) detailed information on the type of buildings in which these individuals live. With this data we can generate not only a synthetic population but also a synthetic building stock. The microcensus records the construction year and number of dwelling units of buildings. This allow us to classify the buildings for the estimation of heat demand.

This procedure has two major advantages: (1) there exist many models for the estimation of heat demand at building level, we can make use of these models, and (2) with the microcensus as the only required data source we are able to estimate heat demand at a spatially disaggregated level for the entire country.

We conclude our paper with an internal validation of the microsimulation model by means of the Total Absolute Error TAE and present the first results from this model aggregated at the NUTS–3 level for the entire country. We briefly discuss the observed patters of the results and attempt to hypothesize on the reasons behind this patterns. We also discuss the difficulties of an external validation of this model and how we can address them in the future.

This article contains 4 figures and 4 tables.

Figures

Figure 1. (a) Distribution of PTAE and (b) difference between estimated and simulated population.
Figure 2. Mean heating consumption per dwelling unit.
Figure 3. Simulated heat demand for Germany at a NUTS-3 level as per capita heat demand in kWh/cap.
Figure 4. Simulated heat demand for selected regions of Germany.

Tables

Table 1.
MC Code* Census Code Unit** Description
EF1 / / Federal State (NUTS 2)
EF952 / Person Weight
EF44 ALTER_KURZ Person Age (five classes of years)
EF49 FAMSTND_AUSF Person Marital status (in detail)
EF46 GESCHLECHT Person Sex
EF20 HHGROESS_KLASS Person Size of private household
EF368 STAATSANGE_KURZ Person Citizenship
EF492 WOHNFLAECHE_20S Dwelling Floor area of the dwelling (20m2 intervals)
EF494 BAUJAHR_MZ Building Year of construction (microcensus classes)
EF635 ZAHLWOHNGN_HHG Building Number of dwellings in a building

Micro Census Code

Refers only to Census

Table 2.
<1859 1860–1918 1919–1948 1949–1957 1958–1968 1969–1978 1979–1983 1984–1994 1995–2001 2002–2009
EFHa 183 180 164 181 146 155 118 132 110 88
RH 153 137 156 106 127 127 98 78 86
KMH 190 143 168 156 129 134 118 122 92 79
GMH 127 144 142 131 117
HH 114 113

source: (Loga et al., 2011) (a) Speci..c Heat demand (spez. Wärmebedarfskennzahl) [kWh/m2a] (EFH) Single family house “Einfamilienhaus”; (RH) Terrace house “Reihenhaus”; (KMH) Apartment house “Mehrfamilienhaus”; (GMH) Large apartment house “Großes Mehrfamilienhaus”; (HH) High-rise “Hochhaus”;

Table 3.
MC Code [S] Code [B] Categories
EF44 0–99 (integers)
ALTER_01JS* (1) Under 18, (2) 18–29, (3) 30–49, (4) 50–64, (5) 65 and over,
S1=∑i=017Si;S2=∑i=1829Si;S3=∑i=3049Si;S4=∑i=5064Si;S5=∑i=65nSi
EF49 (1) Single, (2) Married, (3) Widowed, (4) Divorced, (5) Party to a civil union, (6) Civil partner deceased, (7) Civil union annulled
FAMSTND_AUSF (1) Single, (2) Married, (3) Widowed, (4) Divorced, (5) Party to a civil union, (6) Civil partner deceased, (7) Civil union annulled, (8) No data
B′i=Bi+1 if rand()<Bi÷∑i=17Bi ∀i=1…7 until ∑i=18Bi=∑i=17B′i
EF46 (1)Male, (2) Female
GESCHLECHT (1)Male, (2) Female
EF20 1–62 (integers)
HHGROESS_KLASS (1) 1 person, (2) 2 persons, (3) 3 persons, (4) 4 persons, (5) 5 persons, (6) 6 or more people
S′6=∑i=662Si
EF368 (1) Only German, (2) German and a second citizenship, (3) Abroad
STAATSANGE_KURZ (1) Germany, (2) Abroad
S′1=S1+S2; S′2=S3
EF492 (-1/-5) not applicable
WOHNFLAECHE_20S (1) Under 40, (2) 40–59, (3) 60–79, (4) 80–99, (5) 100–119, (6) 120–139, (7) 140–159, (8) 160–179, (9) 180–199, (10) 200 and more
EF494 (1) Pre–1919, (2) 1919–1948, (3) 1949–1978, (4) 1979–1986, (5) 1987–1990, (6) 1991–2000, (7) 2001–2004, (8) 2005–2008, (9) 2009 and later, (99) no response, (-1/-5) not applicable
BAUJAHR_MZ (1) Pre–1919, (2) 1919–1948, (3) 1949–1978, (4) 1979–1986, (5) 1987–1990, (6) 1991–1995, (7) 1996–2000, (8) 2001–2004, (9) 2005–2008, (10) 2009 and later
B′6=B6+B7
EF635 (1) 1 dwelling, (2) 2 dwellings, (3) 3 dwellings, (4) 4 dwellings, (5) 5 dwellings, (6) 6 dwellings, (7) 7–12 dwellings, (8) 13–20, (9) 21 or more dwellings, (-1/-5) not applicable
ZAHLWOHNGN_HHG (1) 1 dwelling, (2) 2 dwellings, (3) 3—6 dwellings, (4) 7—12 dwellings, (5) 13 or more dwellings
S′3=S3+S4+S5+S6S′4=S7S′5=S8+S9

See Table 1 for a description of the code names

Sum is not equal to total population, an extra category is added to match the total population

Table 4.
BJA # DU SQM BTYP
bja = 1 du = 1 sqm < 600 Btyp03
bja = 1 du > 1 sqm < 600 Btyp04
bja = 1 du > 1 4000 < sqm > 600 Btyp05
bja = 1 du > 1 10000 < sqm < 4000 Btyp06
bja = 2 du = 1 sqm < 600 Btyp07
bja = 2 du > 1 sqm < 600 Btyp08
bja = 2 du > 1 4000 < sqm > 600 Btyp09
bja = 2 du > 1 10000 < sqm < 4000 Btyp10
bja = 3 du = 1 sqm < 600 Btyp15
bja = 3 du > 1 sqm < 600 Btyp16
bja = 3 du > 1 4000 < sqm > 600 Btyp17
bja = 3 du > 1 10000 < sqm < 4000 Btyp18
bja = 3 du > 1 sqm > 10000 Btyp19
bja = 4 du = 1 sqm < 600 Btyp25
bja = 4 du > 1 sqm < 600 Btyp26
bja = 4 du > 1 4000 < sqm > 600 Btyp27
bja = 5 du = 1 sqm < 600 Btyp28
bja = 5 du > 1 sqm < 600 Btyp29
bja = 5 du > 1 4000 < sqm > 600 Btyp30
bja = 6 du = 1 sqm < 600 Btyp31
bja = 6 du > 1 sqm < 600 Btyp32
bja = 6 du > 1 4000 < sqm > 600 Btyp33
bja = 7 du = 1 sqm < 600 Btyp34
bja = 7 du > 1 sqm < 600 Btyp35
bja = 7 du > 1 4000 < sqm > 600 Btyp36