土木工程英文文献及翻译
in Nanjing, China
Zhou Jin, Wu Yezheng *, Yan Gang
Department of Refrigeration and Cryogenic Engineering, School of Energy and Power Engineering, Xi’an Jiao Tong University,
Xi’an 710049, PR China
Received 4 April 2005; accepted 2 October 2005
Available online 1 December 2005
AbstractThe bin method, as one of the well known and simple steady state methods used to predict heating and cooling energy
consumption of buildings, requires reliable and detailed bin data. Since the long term hourly temperature records are not
available in China, there is a lack of bin weather data for study and use. In order to keep the bin method practical in China,
a stochastic model using only the daily maximum and minimum temperatures to generate bin weather data was established
and tested by applying one year of measured hourly ambient temperature data in Nanjing, China. By comparison with the
measured values, the bin weather data generated by the model shows adequate accuracy. This stochastic model can be used
to estimate the bin weather data in areas, especially in China, where the long term hourly temperature records are missing
or not available.
Ó 2005 Elsevier Ltd. All rights reserved.
Keywords: Energy analysis; Stochastic method; Bin data; China
1. Introduction
In the sense of minimizing the life cycle cost of a building, energy analysis plays an important role in devel-
oping an optimum and cost effective design of a heating or an air conditioning system for a building. Several
models are available for estimating energy use in buildings. These models range from simple steady state mod-
els to comprehensive dynamic simulation procedures.
Today, several computer programs, in which the influence of many parameters that are mainly functions
of time are taken into consideration, are available for simulating both buildings and systems and performing
hour by hour energy calculations using hourly weather data. DOE-2, BLAST and TRNSYS are such
* Corresponding author. Tel.: +86 29 8266 8738; fax: +86 29 8266 8725.
E-mail address: yzwu@mail.xjtu.edu.cn (W. Yezheng).
0196-8904/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2005.10.006
Nomenclature
Z. Jin et al. / Energy Conversion and Management 47 (2006) 1843–1850
number of days
frequency of normalized hourly ambient temperature
MAPE mean absolute percentage error (%)
number of subintervals into which the interval [0, 1] was equally divided
number of normalized temperatures that fall in subinterval
probability density
hourly ambient temperature (°C)
normalized hourly ambient temperature (dimensionless)
weighting factor
Subscripts
calculated value
measured value
max daily maximum
min daily minimum
programs that have gained widespread acceptance as reliable estimation tools. Unfortunately, along with
the increased sophistication of these models, they have also become very complex and tedious to
use [1].
The steady state methods, which are also called single measure methods, require less data and provide
adequate results for simple systems and applications. These methods are appropriate if the utilization of
the building can be considered constant. Among these methods are the degree day and bin data methods.
The degree-day methods are the best known and the simplest methods among the steady state models.
Traditionally, the degree-day method is based on the assumption that on a long term average, the solar
and internal gains will offset the heat loss when the mean daily outdoor temperature is 18.3 °C and that
the energy consumption will be proportional to the difference between 18.3 °C and the mean daily tempera-
ture. The degree-day method can estimate energy consumption very accurately if the building use and the
efficiency of the HVAC equipment are sufficiently constant. However, for many applications, at least one
of the above parameters varies with time. For instance, the efficiency of a heat pump system and HVAC equip-
ment may be affected directly or indirectly by outdoor temperature. In such cases, the bin method can yield
good results for the annual energy consumption if different temperature intervals and time periods are
evaluated separately. In the bin method, the energy consumption is calculated for several values of the outdoor
temperature and multiplied by the number of hours in the temperature interval (bin) centered around that
temperature. Bin data is defined as the number of hours that the ambient temperature was in each of a set
of equally sized intervals of ambient temperature.
In the United States, the necessary bin weather data are available in the literature [2,3]. Some researchers
[4–8] have developed bin weather data for other regions of the world. However, there is a lack of information
in the ASHRAE handbooks concerning the bin weather data required to perform energy calculations in build-
ings in China. The practice of analysis of weather data for the design of HVAC systems and energy consump-
tion predictions in China is quite new. For a long time, only the daily value of meteorological elements, such as
daily maximum, minimum and average temperature, was recorded and available in most meteorological
observations in China, but what was needed to obtain the bin weather data, such as temperature bin data,
were the long term hourly values of air temperature. The study of bin weather data is very limited in China.
Only a few attempts [9,10] in which bin weather data for several cities was given have been found in China.
Obviously, this cannot meet the need for actual use and research. So, there is an urgent need for developing bin
weather data in China. The objective of this paper, therefore, is to study the hourly measured air temperature
distribution and then to establish a model to generate bin weather data for the long term daily temperature
data.2. Data used
Z. Jin et al. / Energy Conversion and Management 47 (2006) 1843–1850
In this paper, to study the hourly ambient temperature variation and to establish and evaluate the model, a
one year long hourly ambient temperature record for Nanjing in 2002 was used in the study. These data are
taken from the Climatological Center of Lukou Airport in Nanjing, which is located in the southeast of China
(latitude 32.0°N, longitude 118.8°E, altitude 9 m).
In addition, in order to create the bin weather data for Nanjing, typical weather year data was needed.
Based on the long term meteorological data from 1961 to 1989 obtained from the China Meteorological
Administration, the typical weather year data for most cities in China has been studied in our former research
[11] by means of the TMY (Typical Meteorological Year) method. The typical weather year for Nanjing is
shown in Table 1. As only daily values of the meteorological elements were recorded and available in China,
the data contained in the typical weather year data was also only daily values. In this study, the daily maxi-
mum and minimum ambient temperature in the typical weather year data for Nanjing was used.
3. Stochastic model to generate bin data
Traditionally, the generation of bin weather data needs long term hourly ambient temperature records.
However, in the generation, the time information, namely the exact time that such a temperature occurred
in a day, was omitted, and only the numerical value of the temperature was used. So, the value of each hourly
ambient temperature can be treated as the independent random variable, and its distribution within the daily
temperature range can be analyzed by means of probability theory.
3.1. Probability distribution of normalized hourly ambient temperature
Since the daily maximum and minimum temperatures and temperature range varied day by day, the con-
cept of normalized hourly ambient temperature should be introduced to transform the hourly temperatures in
each day into a uniform scale. The new variable, normalized hourly ambient temperature is defined by
^ ¼ttmintmaxtmin
where ^ may be termed the normalized hourly ambient temperature, tmaxand tminare the daily maximum and
minimum temperatures, respectively, t is the hourly ambient temperature.
Obviously, the normalized hourly ambient temperature ^ is a random variable that lies in the interval [0, 1].
To analyze its distribution, the interval [0, 1] can be divided equally into several subintervals, and by means of
the histogram method [12]i
in each subinterval can be calculated by1137