Esp-r Software Free Download
This paper describes the current status of the ESP-r simulation environment with respect to modelling and simulation of systems in buildings. Following an overview of the underlying philosophy and the overall structure of the system, ESP-r's integrated approach to building and plant modelling is elaborated with a worked example. The paper finishes with indicating current system simulation developments and concludes that an integral approach to a building with associated systems is now possible and ready to be applied on a wide scale in engineering education and research, as well as in practice. 1. INTRODUCTION Apart from the general need for energy efficiency and protection of the environment, building designers and environmental engineers encounter a vast range of additional "problems" in everyday practice. To name just a few examples: in offices phenomena like "Sick Building Syndrome" and trends like: atria, climate facades, displacement ventilation combined with cooled ceilings; in...
Figures - uploaded by Jan LM Hensen
Author content
All figure content in this area was uploaded by Jan LM Hensen
Content may be subject to copyright.
Discover the world's research
- 20+ million members
- 135+ million publications
- 700k+ research projects
Join for free
Current Building Systems Modelling Potential of ESP-r
E O Aasem, J A Clarke, J L M Hensen, N J Kelly, J MacQueen, C O R Negrao
Energy Simulation Research Unit
University of Strathclyde, GLASGOW G1 1XJ, UK
tel: +44 41 552 4400 fax: +44 41 552 8513 email: esru@esru.strathclyde.ac.uk
ABSTRACT
This paper describes the current status of the ESP-r simulation environment with
respect to modelling and simulation of systems in buildings. Following an overview of
the underlying philosophy and the overall structure of the system, ESP-r's integrated
approach to building and plant modelling is elaborated with a worked example.
The paper finishes with indicating current system simulation developments and
concludes that an integral approach to a building with associated systems is now
possible and ready to be applied on a wide scale in engineering education and research,
as well as in practice.
1. INTRODUCTION
Apart from the general need for energy efficiency and protection of the environment, building designers
and environmental engineers encounter a vast range of additional "problems" in everyday practice. To
name just a few examples: in offices phenomena like "Sick Building Syndrome" and trends like: atria,
climate facades, displacement ventilation combined with cooled ceilings; in houses increasing levels of
occupants comfort expectancy; in industry (indoor) air quality issues; in conversions where existing
buildings have to house completely different functions (eg industrial buildings converted into
apartments); etc.
So which is the system we are actually trying to address? The whole of building form and fabric, control
systems, environmental issues and methodical design comprises a very wide area. Many of the above
indicated problems are in fact caused by the complexity due to interactions between the various sub-
fields. These interactions are indicated in Figure 1. Obviously this diagram is merely a gross
simplification of reality, because in the real world this is a n-dimensional problem involving the
3-dimensionality of building and plant, the dimension of time, and the dimension of the various aspects
like: thermal environment, air quality, lighting, acoustics, etc.
As illustrated for the thermal aspects only, the indoor environment is determined by a number of sources
acting via various heat and mass transfer paths. The main sources may be identified as outdoor climate,
occupants (casual heat gains), and the auxiliary system which may perform heating, ventilating and / or
air-conditioning (HVAC) duties. These sources act upon the indoor environment via various heat and
mass transfer processes such as conduction, solar transmission, long wave radiation exchange,
convection, airflow, and flow of fluids within the plant system. Within the overall configuration as
sketched, several energy sub-systems may be identified, each with their own dynamic thermal
characteristics: occupants (very complicated dynamic systems themselves), building structure, and
auxiliary systems. The cycle periods of the excitations acting upon the system are also highly diverse.
They range from something in the order of seconds for the plant, via say minutes in case of the occupants,
-1-
gy g y,
Figure 1 The building as an integrated, dynamic system
to hours, days and year for the outdoor climate.
Many design problems (for example fabric design, comfort or condensation assessment, control system
appraisal, plant system analysis, etc.) can only be meaningfully assessed when treated as a sub-set of
some complex set of interactions. In other words, a piecemeal approach, in which a particular region is
considered in isolation, is often inappropriate and potentially misleading.
Having identified the problem domain and the need for an integral approach, we are now able to state our
objective: evaluation of building performance while treating the building (including its distributed flow
paths), environmental control system(s), and occupants as an integrated, dynamic system.
The remainder of this paper dwells on one particular simulation system, enabling the above, by first
outlining the system and then elaborating the integral approach by means of a worked example.
2. OUTLINE of ESP-r
ESP-r is a building energy modelling environment which supports performance assessment of design
solutions (for either newly build or alterations of existing stock) incorporating traditional and/or low
energy features. The approach is intended to allow users to conduct a high integrity, first principle
appraisal whilst modelling all aspects of the energy subsystems simultaneously and in the transient
domain. In common with other simulation programs, the ESP-r approach is markedly different from
traditional methods in that it aims to represent all relevant phenomena, and to process these phenomena
simultaneously so that the inter-relationships are preserved. Essentially, this is achieved by establishing
sets of conservation equations for different spatial regions and arranging for the integration of these
equations over time. In this way the energy and mass flows are tracked - throughout a simulation - as they
evolve under the influence of climatic boundary conditions, occupancy effects inside the building,
constraints imposed by any control action and by the potentially time-dependent inter-volume links
(representing for example damper, valve, window or door movement).
The theories employed by ESP-r to represent heat transfer and fluid flow, and the numerical techniques
used to achieve equation integration, are detailed elsewhere (Clarke 1985, Hensen 1991).
Figure 2 summarizes the ESP-r system. Essentially, the system consists of two distinct parts
corresponding to the user's and researcher's viewpoints.
Typically, users require considerable assistance with the specification of a design hypothesis and its
evolution in the light of poor performance indications. These functions are provided by a Project
Manager which supports the specification of design problems in terms of:
-2-
gy g y,
Figure 2 The ESP-r system: user and researcher viewpoints
• 3D building geometry with attribution in relation to opaque and transparent constructional materials,
surface finishes, occupancy, lighting schemes, air leakage distribution, with superimposed events to
represent phenomena such as window opening, shading device positioning and electric light
switching.
• Plant specification in terms of networks of connected components representing the thermodynamic
processes which result in the pressure and temperature differences causing heat transfer between,
and the flow of, the working fluids.
• Control system specification in terms of a list of control loops, possibly nested, constructed from
sensor->action->actuator relationships, each one valid over a given time interval.
When specifying a problem, the Project Manager offers users access to on-line databases of
constructional materials, plant components, profile prototypes, optical properties, pressure coefficients
and climatic sequences of differing severity. In this way the input burden is minimized. Where possible,
inputs are achieved through graphical interaction, with customized tools invoked automatically by the
Project Manager. For example, AutoCad can be used to specify building geometry, while an icon
manipulation program has recently been added to assist with the definition of plant and flow networks.
Researchers, on the other hand, are usually concerned to extend modelling functionality and accuracy by
experimenting with alternative algorithms. This requires a mechanism for working with source code
which does not oblige the researcher to be knowledgeable about all aspects of the system. With ESP-r this
has been achieved by compartmentalizing the source code into technical domains as shown in Figure 2. A
researcher interested in plant modelling, for example, need only work with the source code relating to that
domain. Software engineering tools and a good practice procedure for software development (Hensen
1991a) are then employed to control the integration of the different domains into the ESP-r Simulator. At
the present time this mechanism is supporting an expansion in the number of organizations actively
developing ESP-r. (The system is currently used for education and research at universities and research
centres in over 20 countries world-wide.)
-3-
gy g y,
Starting from such an existing platform offers vast advantages for any individual research group. The
most important ones are: (1) as an individual group it is not necessary to have expertise in all areas, (2)
areas not addressed within a specific research project will still be state-of-the-art, (3) as more people are
using the system, any bugs or flaws are likely to surface sooner, (4) common tasks like development of
training/ tutorial material becomes a shared burden, and (5) results transfer to the international research
community is implicit and therefore very efficient.
3. A WORKED EXAMPLE
Application of this integral simulation system can best be demonstrated by means of a realistic case study.
The case presented here concerns the modelling a small building at a site near London. Air conditioning
is achieved with a single-zone system with re-circulation. A control strategy has been suggested to keep
the air temperature and relative humidity of the main zone within certain limits. Simulation is employed
to assess this proposed control strategy.
Building
Figure 3 Definition of the building model
The building model comprises three zones, each representing a distinct area of the building. The three
zones being: display area, storage, and roofspace. The building is shown in Figure 3. The display area is
occupied during normal working hours (9 - 17), with the resulting latent and sensible energy gains due to
people and equipment described within the problem description files. To keep the problem relatively
simple, time dependent air infiltration and inter-zonal ventilation flow have been pre-defined in this case.
Alternatively we might have defined the building's air leakage distribution in terms of cracks, vents,
-4-
gy g y,
doors, openings and external surface wind pressure coefficients. In that case the mass flow simulation
module would have solved - during each simulation time-step - the airflow due to wind and temperature
effects (as elaborated in Hensen and Clarke 1991).
At simulation time, the building so defined is made discrete by subdivision into a number of
interconnecting, finite volumes. These volumes then possess uniform properties which can vary over
time, and represent homogeneous and mixed material regions associated with room air, room surface and
constructional elements. For this particular problem the number of such volumes is about 190. Then, for
each of these finite volumes in turn, and in terms of all surrounding volumes deemed to be in thermal or
flow contact, a conservation equation is developed in relation to the transport properties of interest - heat
energy or mass exchange for example. This gives rise to a whole system equation-set where each equation
represent the state of one finite volume as it evolves over some small interval of time. Once established
for a particular increment in time, the equation-set is simultaneously solved - by a numerical method -
before being re-established for the next time-step of some user-specified simulation period. In support of
equation-set generation, many algorithms are required to compute such information as solar and casual
gains, sky and ground temperatures, heat transfer coefficients and so on.
Plant
In keeping with early design investigations ESP-r provides a level of plant systems which are 'ideal' in
their representations ie. they have no inertia or time dependent characteristics, and operate on the building
side only.
Figure 4 Diagrammatic representation of solving the plant system matrix for energy
balance (ISTAT = 1), 1st phase mass balance (ISTAT = 2), and 2nd phase mass balance
(ISTAT = 3); see Table 1 for brief explanation of various subroutines; IFLWN indicates
whether the mass flow network solver is active.
However the system also provides a much more detailed level of plant simulation, using a modular-
simultaneous technique. In this case system plant system modelling is achieved by a modular, component-
wise approach. Each plant component is made discrete by subdivision into one or more interconnected
finite volumes. For each finite volume (or node) up to three conservation equations are developed to
-5-
gy g y,
represent heat and mass transfer. In case a node represents a solid region there is only one equation (heat),
for a node representing water there are two equations (heat + water flow), and for a node representing air
there will be three equations (heat + dry air + water vapour).
The plant model is a combination of individual component models forming up to three complete sets of
simultaneous state-equations for the whole system, for which a solution is found by means of efficient
matrix equation solvers and a procedure as indicated in Figure 4 and Table 1.
Table 1 Main subroutines involved in solving plant system matrices
Subroutine Description
MZPMXT accesses the plant components database and extracts the data needed to establish the templates for
the plant network energy, 1st phase mass, and 2nd phase mass balance. Checks that component
connections are legally defined, initialises data arrays, and in case the encapsulated version of mfs
is active: resets the mass diversion ratios to unity, and checks plant/mfs connection mapping
regarding fluid types
MZPMRX controls setting up, solution and results assignment of the plant energy (ISTAT = 1), 1st phase mass
(ISTAT = 2), and 2nd phase mass (ISTAT = 3) matrix equation at each plant simulation time step
MZPADJ organises the information which defines each inter-component coupling. This data is required by
the component nodal equation coefficient generators in order to calculate the inter-component
connection coefficients. Checks whether mass flows are in the assumed direction (ie. the coefficient
generators implicitly assume each connection's mass flow rate
≥ 0. Establishes plant component
containment temperatures, if defined to exist
CONTRL determines plant control status based on most recent available results, by invoking appropriate
control routine for each active plant control loop for current time step
MFLWCA controls the fluid flows calculation for each simulation time step: sets climate variables,
temperatures for nodes corresponding to plant components or building zones, sets boundary nodes
temperature and/or wind pressure. Calculates fluid densities and connections stack pressure
difference. Solves the fluid flow network mass balances; transfers fluid flow simulation results to
results file. Establishes and transfers building side air flow and plant side fluid flows
MZPMSU sets up the plant matrix equations by calling the appropriate matrix coefficient generators, and
locating generated equation coefficients in the network matrix
MZPMSV solves the plant matrix equation. A sparse matrix solver is invoked to solve the matrix equation
A
.
θ
= b or E
.
.
m = f for the solution vector
θ
or
.
m in terms of the known vector b or f. The solution
vector is then relocated in the future time-row state-space variable array.
MZNASS adjusts all plant related history variables
For the current case-study it was decided to use one of the most common and effective approaches to
controlling temperature and relative humidity within a zone: pre-heat the air, pass it through an air washer
where it undergoes adiabatic saturation, and then to re-heat it to the temperature at which it is to be
supplied to the zone. The pre-heating and adiabatic saturation processes will permit the relative humidity
in the zone to be controlled and re-heating allows the temperature therein to be properly regulated during
winter conditions.
The system network for the current case consists of 14 components describing the major items of plant of
a typical year-round, single-zone system. The components required to create the system are: pre-heat and
re-heat coils, cooling coil, supply and return fan, air washer, mixing box, and ducts. The models which
were used in this case all originate from the IEA Annex 10 work (1988). The system layout is shown in
Figure 5; the corresponding energy balance matrix is shown in Figure 6. For the current case-study, the
mass flow balance matrices for dry air and water vapour will have a similar lay-out, but obviously
different coefficients and right hand side values.
The pre-heat and re-heat coils are one-node flux control heating coils, the maximum output of each coil is
3.5
kW . The cooling coil is also a relatively simple flux control model with a maximum capacity of 1
kW . If a greater level of granularity is required in the simulation the described components can be
replaced by the more detailed 3-node models also available in the plant component database, where a
-6-
gy g y,
Figure 5 Definition of the air-conditioning system
Figure 6 Energy equations matrix lay-out for the plant system
-7-
gy g y,
fluid flow rate to the heater/ chiller unit is controlled as opposed to flux.
The air washer (or spray humidifier) in the system can either be uncontrolled (ie fixed moisture addition)
or be controlled to vary the water flow rate according to a certain strategy.
The intake and extract in the network provide a constant volume flow rate of 0.4
m
3
/ s to and from the
conditioned space; this corresponds to an air change rate of about 5 ACH . The supply air has a constant
fresh air content of 20%, with the remainder made up from re-circulated zone air.
Similar to the remark above regarding building side airflow due to infiltration and ventilation,
alternatively we might have decided to define the flow behaviour of the system in terms of pressure-flow
characteristics by selecting appropriate models for the various plant components (ie fan, mixing box, duct,
etc). In that case the mass flow simulation module would have solved - during each simulation time-step -
the airflow through the system.
Controls
In ESP-r a control loop is defined by: sensor location, actuator location, and control law. A control loop is
not static but has a temporal dimension so that it can change as a function of time. A sensor exists to
measure some variable for transmission to the control law representing the active controller. The control
variable may be any nodal state variable active within a simulation, an outdoor condition, one of the plant
component additional variables, or some derived combination of the previous. Actuators exist to transmit
the output of a controller to some building zone or plant component, usually to reduce the deviation of the
sensed control variable from some user-specified set point. Actuator locations can be set to any building
side node (air, surface, mixed, intra-construction), or some plant component participating in a simulation.
A control law is an algorithm which represents the logic of some controller. Its purpose is to translate
(algorithmically) the sensed condition to the actuated state in terms of the control system characteristics,
for example: building pre-heat, fixed heat injection or extraction, PID control, optimum start controller,
etc. (For each control loop, in addition to specifying sensor, actuator, and control law data, the user has to
specify items such as: periods of validity and operation, component output capacity, set point, throttling
range, etc,.)
At each time-step as a simulation proceeds, the nodal property detected by the sensor is fed to the control
law algorithm, which then acts to fix or limit some other nodal property, via the actuator node, prior to
matrix reformulation for the current time-step. In this way, simulation control is achieved on the basis of
some function of a prevailing control point.
At an early design stage it is useful to be able to conduct simulations on the basis of the assumption of
ideal control. This allows factors such as energy efficiency to be improved by the systematic adjustment
of the building design parameters against the expectation of ideal comfort conditions. Subsequently, after
a near optimum design has been arrived at, more realistic control regimes can be imposed in order to
obtain a stable control system with good response characteristics. The latter is being pursued in the
current case-study.
The determination of supply conditions for such a system requires the determination of values for the dry-
bulb temperature and moisture content of the supply air which are required to maintain a design state in a
conditioned space. The control law used in the current case-study is a psychometrically-based algorithm
which establishes and re-sets (at every plant time-step) the set-points for the pre-heater, re-heater, cooling
coil, and humidifier plant components. For the current system four control loops (as detailed in Table 2)
are active: one each for the pre-heater, cooling coil, humidifier, and re-heater components.
Coupling Building and Plant
In a mathematical/ numerical sense, this effectively means combining the energy and flow balance matrix
equations for both the building and plant. While in principle it is possible to combine all building / plant
and heat / fluid flow matrix equations into one overall 'super-matrix', this is not done primarily because of
-8-
gy g y,
Table 2 Plant control loop parameters for pre-heater (1), cooling coil (2), humidifier (3)
and re-heater (4); set-points for room conditions: 19
°C dry bulb temperature and 50%
RH.
Description (1) (2) (3) (4)
Sensor Location: pre-heater exit coil exit humidifier exit re-heater exit
Sensed Variable: dry bulb temp. dry bulb temp. relative hum. dry bulb temp.
Actuator Location: pre-heater cooling coil humidifier re-heater
Actuated Variable: heating flux cooling flux moisture injection heating flux
Control Law: proportional proportional proportional proportional
Proportional Band: 2
°C 2 ° C 12% RH 2 °C
Output Range: 0.0 -> 3.5 kW 0.0 -> 3.5 kW 0.001 -> 0.005 kg / kg
da
0.0 -> 3.5 kW
Control Period: 07:00 - 18:00 07:00 - 18:00 07:00 - 18:00 07:00 - 18:00
the advantages which accrue from problem partitioning.
The most immediate advantage is the marked reduction in matrix dimensions and degree of sparsity. A
second advantage is that it is possible to easily remove partitions as a function of the problem in hand; for
example when the problem incorporates building only considerations, plant only considerations, plant +
flow, and so on. A third advantage is that different partition solvers can be used which are well adapted
for the equation types in question - highly non-linear, differential and so on.
Obviously there are often dominating thermodynamic and/ or hydraulic couplings between the different
partitions. If a variable in one partition (say air temperature of a building zone) depends on a variable of
state solved within another partition (say the temperature in the air supply), it is important to ensure that
both values match in order to preserve the thermodynamic integrity of the system. Without going into
details, two methods are offered to handle these couplings: (1) a time step control facility, and (2)
iteration mechanisms.
Figure 7 Indicative flow chart showing the main loops in the simulation process for a
combined building and plant configuration
-9-
gy g y,
Figure 7 visualizes ESP-r's main numerical controller (MZNUMA) which controls the simulation process
for combined building and plant configurations. As indicated in this diagram, the overall configuration
simulation time increments may be smaller than one hour.
†
A complete configuration time step involves
the evaluation of all building-side zones followed by the processing of the plant system equations. If a
mass flow network is defined to exist, this is processed together with the plant system network. In case the
user defined a building-only configuration, the mass flow network is processed prior to building zones.
At each overall configuration simulation time step the building- and plant-side state-space equations, and
the mass flow network equations are generated and solved from up to five separate matrix equations. The
building-side solution process is invoked once per user-specified time step. This process uses a matrix
partitioning technique (ie. one partition for each building zone) as described by Clarke (1985). For the
building, heat input or extraction by the plant are regarded as as known boundary conditions.
Since it is practice to process the plant equations at a greater frequency than building matrices (because of
the different time constants), the plant matrix may be established at some sub-interval of the building time
step. For the plant, the connections with the building are treated as excitations. Then the plant matrix is
solved by a sparse matrix method as described above.
Division of the overall simulation problem in a building-side and a plant-side may leed to certain
difficulties. When processing the building-side energy balance, heat input or heat extraction by the plant
for the time step under consideration should be known. It is common practice to use plant side
temperatures and mass flow rates from the previous time step in evaluating this heat exchange. When
building-side control is based on a plant-side originating signal a similar time shift occurs. When
processing the plant-side energy balance, the component losses are calculated with containment (perhaps
building-side) temperatures which were calculated with plant-side state variable values from the previous
plant time step. A similar effect may occur when plant-side control is based on a signal originating from
the building-side.
One way to deal with this kind of problems, is to make use of a mechanism such as indicated in Figure 7
which could be labeled as a mixed direct/iterative solution scheme. At the indicated point in the
calculation process, the plant heat input as assumed in processing the building side is compared with the
plant heat emission as calculated when processing the plant side. If the difference exceeds some user
specified value, the whole building and plant solution process is repeated based on the newly calculated
values. If either the absolute or the relative difference between assumed and newly calculated
building/plant heat exchange satisfies the user specified tolerances, the model proceeds with the next time
step. In order to prohibit excessive number of iterations, the iteration process may only be enabled when
the user specifies one plant time step per building time step.
Simulation
As an example of the operation and interaction of the plant and building models a few typical days in
January are simulated. The climate data used represents a typical year for the South of England.
In winter psychrometric processes the active components in the plant network are the pre-heat, re-heat
and humidifier.
Pre-heating serves two functions, the first is that varying the output of the heater allows control of the
amount of moisture evaporated into the process air at the humidifier. The second function performed by
the pre-heater is to prevent freezing of water in the humidifier.
Variation of re-heater output allows control of the supply air dry bulb temperature. The humidification
process serves to offset the low moisture content of the outside air.
The display area is the focus of the simulation, with the plant attempting to control the zone dry-bulb
temperature to 19
°C and a relative humidity of 50%.
†
Note that by choosing a time step the user implicitly decides to ignore the process dynamics within the
time step
-10-
gy g y,
The simulation is run over the period 7-9 January. The time-step used on the building side is 15 minutes,
while on the plant side the time-step is 1 minute. The time-step is kept small to capture the dynamic
performance of the plant components, and also to avoid iteration problems which occur with larger plant
time-steps. It should be noted that this has nothing to do with the solution process itself, but is a
consequence of how the problem is defined; ie here we use flux controlled coils which means that
normally the heat input or extraction will either be maximum or minimum. Now if the time-step is too
large, this will cause the temperature rise or drop to be too large, resulting in oscillating behaviour.
Results
Figure 8 Results analysis module showing the zone conditions
Some of the results of the simulation process are shown in Figure 8. It can be seen from the results that
the control system performance for relative humidity within the zone is good throughout the simulation
period. Relative humidity is held close to the prescribed set-point of 50%. However the dry bulb
temperature is consistently higher than the set-point of 19
°C for prolonged periods during the day. This
can be attributed to poor controller tuning. Clearly, controller tuning parameters could be altered to
reduce further such factors as overshoot and deviation from set-point.
It should be noted that - in the current case-study - the capacities for each actuated component are fixed
for the duration of any particular control period. These could be altered in a dynamic manner by means of
some (weather) compensating control loop. Also, controllers often tend to act in a 'conflicting' manner,
e.g. cooling and humidification control is one such example. For such cases, a 'control supervisor',
facility is required to resolve such conflicts. (A model for such a facility is currently under development.)
Although the results might be interpreted as poor control behaviour, it is also important to note that this is
-11-
gy g y,
due to the performance of the "real controller" rather than of the simulation tool. This example illustrates
that simulation can be utilised to identify poor system performance.
Figure 9 Results analysis module showing the psychrometric processes occurring within
the plant.
Figure 9 shows the various psychrometric processes occurring within the plant when the controls are
active. These processes include mixing between fresh and return air, humidification, and re-heat. The
temperature of the binary mixture entering the pre-heat coil is such that pre-heating is not required in this
instance. A slight amount of cooling is required to off-set a degree of over-humidification in the washer.
However this could be overcome through modification of the psychrometric control algorithm such that it
will allow tighter control of the humidification process.
4. CURRENT DEVELOPMENTS
As indicated in the introduction, many researchers in various locations are now working with the ESP-r
environment. In most cases this concerns application and validation of the system, but in an increasing
number of cases this also involves new developments. Sometimes this is adding a small feature or
improving certain aspects of the system, in other cases it involves completely new modules. Often this
work is done in the context of post-graduate studies.
Current developments introducing adaptive gridding techniques (enabling explicit modelling of three
dimensional phenomena such as thermal bridging and constructional edge effects), variable material
properties, modelling of combined heat and moisture transfer, artificial/ˆdaylight modelling, and design
tool cooperation (eventually leading to integrated building design systems) are not covered in this paper.
-12-
gy g y,
Here we just want to indicate some developments which most pertain to system simulation in buildings.
Conjoining CFD and Building Simulation
This project (Negrao 1994) has implemented within ESP-r a CFD module capable of solving dynamic
three-dimensional turbulent airflow and buoyancy effects. The model uses the widely used
k −
ε
model
for representing turbulence, the SIMPLE algorithm (Patankar 1980) for linking pressures with velocities,
a staggered (possibly non-uniform) grid, and an up-wind scheme to approximate the density in the
velocity cells.
The CFD module is used to obtain the mass, energy and momentum state of a space when connected to a
nodal network representing the distributed flowpaths throughout the building, its plant and systems. In
this way the entire building system can be processed simultaneously, with increased resolution within user
nominated spaces.
Plant Simulation
A continuing effort is expanding the existing plant components database. Developing or adjusting models
is a difficult and time-consuming process which can be alleviated by reusing existing models. In this
context ESP-r now enables incorporation of "external" models like for instance TRNSYS type models.
Another mechanism is by creating a "neutral model format"-translator which will enable automatic
incorporation of models written in NMF (Bring et al 1992).
In terms of the plant components database recently additional models, based on energy and mass
conservation, have been added for a range of components as found in air conditioning systems (Aasem
1993), and in solar systems. In addition work has begun (Kelly 1994) on the modelling of combined heat
and power (CHP) units. These component models can then be connected to define a plant system and
subjected to a dynamic analysis simultaneously with the building.
Near future efforts will concentrate on the development of a power module to allow simulation of
electrical loads and sources within the building model. This development, as well as augmenting plant
and CHP models will also allow the simulation of building-oriented self-contained power systems
containing wind, photovoltaics, power storage and other electrical power sources.
Another project (Chow 1993) is attempting to establish mathematical models for each of the physical
processes that occur within plant components (boiling heat transfer, flame radiation, fluid flow, etc.) and
to use these to explore the possibility of automatically constructing component models from primitive
parts. If successful, this will allow all component models to be synthesized from a small number of
primitive models rather than, as at present, each component requiring a unique mathematical model. In
this way, it will be possible to create a simulation system which is plant-type independent while being
able to handle all plant types: just as contemporary building simulators are essentially building type
independent.
Control System Simulation
The objective of this project (MacQueen 1993) is to extend ESP-r's family of controllers so that they
cover a broad spectrum from simple systems, such as thermostatic devices, to state-of-the-art Building
Energy Management Systems (BEMS) capable of reacting to changes in both the external and internal
environments associated with a building in a fully integrated way. This entails the development of new
sensor and actuator models capable of multiple variable action, control algorithms to represent advanced
sensor-to-actuator relationships and control executives to coordinate the above against the user specified
control strategy.
Recent developments in terms of control involve both "traditional" controllers like PID, duty cycle, and
cascade controllers, but also novel controller types like fuzzy logic rules-based and the use of ESP-r as a
simulation-based logic controller (where ESP-r is the process model which is used to forecast future
reality, and the controller algorithm is actually a set of logic statements). In terms of application the latter
-13-
gy g y,
facility can be used as a design tool, a commissioning aid, a fully adaptive controller with its parameters
re-evaluated at every time-step, or an on-line BEMS controller (for example optimum start/stop controller
or adaptive PID controller).
Validation
Significant effort is being expended, particularly in Europe, on the topic of model validation. In recent
times, major contributions to the development of a coherent validation methodology have been made by
for instance the International Energy Agency (Judkoff 1994) and the Commission of European
Communities (CEC 1993). Because of these efforts a consensus has emerged on the components of such
a methodology: review of theory and algorithms, code checking, analytical testing, inter-model (or
program) comparison, sensitivity analysis, and empirical testing.
The next important step is to organize the methodology so that it can be routinely applied in practice.
Within the ESP-r environment this goal is being pursued by the implementation of several validation
"aids" like: CASE tools to help detect/eliminate programming errors; analytical tests which can be
invoked whenever the code is upgraded; a series of on-line benchmark tests suitable for inter-program
comparison; test cases for which substantial monitored data exists, and supporting the statistical
comparison of predicted and monitored data; and last but not least improving user feedback through
improved user interfaces and more extensive consistency and range checking.
5. CONCLUSIONS
This paper has described and demonstrated an approach for the simulation of combined heat and fluid
flow in a building / plant context. Typically a model's focus is either the building or the plant side. Here
we have a simulation environment of combined potential, which attempts to process the entire system in
the dynamic state and to the same level of detail. The present performance of the model indicates that it is
now feasible and practical to solve complex building/ plant/ flow networks on currents small computer
systems. This enables an integral approach of the thermal interactions between a building, its
environmental control system, the occupants, and the context.
Computer simulation using advanced and state-of-the-art models, is quickly becoming accessible to the
(environmental) engineering community. It is now ready to be applied on a wide scale in engineering
education and research, as well as in practice.
Acknowledgments
The results achieved thus far, would not have been accomplished without the continuing support of many
people. For this the authors wishes to express their sincere gratitude to all ESP-r colleagues.
References
Aasem, E.O. 1993. ''Practical simulation of buildings and air-conditioning systems in the transient
domain,'' PhD thesis University of Strathclyde, Glasgow.
Bring, A., P. Sahlin, and E. Sowell 1992. ''The Neutral Model Format for building simulation,''
Installationsteknik Bulletin no 24, Royal Institute of Technology, Dept. of Building Services
Engineering, Stockholm.
CEC 1993. The PASSYS Project: Subgroup Model Validation and Development - Final Report
1986-1992, Commission of the European Communities, DG XII of Science, Research and
Development, Brussels. S. O/ stergaard Jensen (Ed.)
Chow, T.T. 1993. ''PhD progress report,'' ESRU Technical Report, University of Strathclyde, Glasgow.
-14-
gy g y,
Clarke, J.A. 1985. Energy simulation in building design, Adam Hilger Ltd, Bristol (UK).
Hensen, J.L.M. and J.A. Clarke 1991. ''A fluid flow network solver for integrated building and plant
energy simulation,'' in Proc. 3rd Int. Conf. on System Simulation in Buildings, Dec 1990,pp.
151-167, University of Liege.
Hensen, J.L.M. 1991. ''Good practice guide for ESP
R
developers,'' Collaborative FAGO/ESRU report
91.38.K, pp. 1-36, Eindhoven University of Technology.
Hensen, J.L.M. 1991. ''On the thermal interaction of building structure and heating and ventilating
system,'' Doctoral dissertation Eindhoven University of Technology (FAGO).
IEA 1988. ''HVAC component specification booklets,'' Energy conservation in buildings & community
systems programme; Final report Annex X, International Energy Agency. Operating agent:
University of Lie`ge, J. Lebrun
Judkoff, R.D. 1994. ''A testing and diagnostic procedure for building energy simulation programs,'' in
Proc. BEP '94 Conference "Facing the Future", York, April 6-8, Building Environmental
Performance Analysis Club (BEPAC), York (UK).
Kelly, N. J. 1994. ''PhD progress report,'' ESRU Technical Report, University of Strathclyde, Glasgow.
MacQueen, J. 1993. ''PhD progress report,'' ESRU Technical Report, University of Strathclyde,
Glasgow.
Negrao, C. 1994. ''PhD progress report,'' ESRU Technical Report, University of Strathclyde, Glasgow.
Patankar, S.V. 1980. ''Numerical heat transfer and fluid flow,'' Series in Computational Methods in
Mechanics and Thermal Sciences, Hemisphere Publ. Corp., New York.
-15-
... ESP-r (Aasem et al. 1994) Enables building performance simulations including heat, air and light. ...
Several approaches exist for simulating building properties (e.g. temperature, noise) and human occupancy (e.g. movement, actions) in an isolated fashion, providing limited ability to represent how environmental features affect human behaviour and vice versa. To systematically model building-occupant interactions, several requirements must be met, including the modelling of (a) interdependent multi-domain phenomena ranging from temperature and sound changes to human movement, (b) high-level occupant planning and low-level steering behaviours, (c) environmental and occupancy phenomena that unfold at different time scales, and (d) multiple strategies to represent occupancy using established models. In this work, we propose an integrated platform that satisfies the aforementioned requirements thus enabling the joint simulation of building-occupant interactions. To this end, we combine the benefits of a model-independent, discrete-event, general-purpose framework with an established crowd simulator. Our platform provides insights on a building's performance while accounting for alternative design features and modelling strategies.
... Focusing first on energy-related software, there are 147 tools currently listed in the Building Energy Software Tools Directory (BEST-D). Many of these tools, however, are based on a few core simulators, such as Radiance (Ward, 1994) for lighting and EnergyPlus (Crawley et al., 2001), DOE-2 (Curtis et al., 1984), or ESP-r (Aasem et al., 1994) for whole building energy simulation. ...
In recent decades, architects have turned to computer simulation with the hope of designing more functional, sustainable, and compelling buildings. In such efforts, it is important to regard buildings not merely as static structures, but rather as complex dynamic systems driven by highly stochastic elements including the weather and human behavior. In this chapter, we describe how simulation has impacted architectural design research and practice. A multitude of simulation tools have been developed to model specific aspects of a building such as thermodynamics, daylight, plug loads, crowd behavior, and structural integrity under internal and external loads. Yet numerous challenges remain. For example, although many factors influencing buildings are interdependent, they are often analyzed in isolation due to the development cost associated with integrating solvers. A systems approach combining visual programming with state-of-the-art modeling and simulation techniques may help architects and building scientists combine their expertise to produce integrated complex systems models supporting emerging paradigms such as generative design.
... Due to space constraints, this needs to be limited to two very brief descriptions. (Other, and more elaborate, case studies may be found in, for example, Aasem et al 1994, andClarke et al 1995.) Although these examples could have been modelled using other building energy simulation environments, the following examples are based on ESP-r. ...
- Dr Ir
-
![Jan LM Hensen]()
This paper attempts to describe the advantages and disadvantages of different modelling approaches for design and performance evaluation of heating, ventilating, and air-conditioning (HVAC) systems for buildings. Merits and drawbacks of the various modelling methods are illustrated by case study material. Finally some conclusions and directions for future work are indicated. Keywords: building energy modelling, building energy simulation, HVAC systems 1. Introduction When speaking about `a building', often, we actually mean the whole of building form and fabric, heating, ventilating, and air-conditioning (HVAC) and other systems. This `whole' comprises a wide area, where many problems occurring in practice are in fact caused by the complexity due to interactions between the various sub-fields. Figure 1 The building as an integrated, dynamic, thermal system Some of the (thermal) interactions are indicated in Figure 1., which is obviously merely a gross simplification of reality, becau...
-
![Jan LM Hensen]()
- J.A. Clarke
Summary This paper aims to outline the current state-of-the-art in integrated building simulation for performance prediction of heating, ventilating and air-conditioning (HVAC) systems. The ESP-r system is used as an example where integrated simulation is a core philosophy behind the development. The current state and future developments are illustrated with case studies. It is argued that for building simulation to penetrate the profession in the near future, there is a need for appropriate training and professional technology transfer initiatives.
HVAC plant configurations vary dramatically in their conceptual design, and in the complex building-and-plant topologies Analysis of plant performance often relies on a powerful and flexible simulation tool to produce realistic results based on appropriate system and component models. The stress is thus on the credibility and applicability of contemporary systems simulation packages and the component libraries that underpin these systems. This paper describes an approach to plant component modelling, which is based on the notion of 'primitive parts' (PP) to represent the fundamental heat and mass transfer processes. The principal advantage of the approach lies in its flexibility in relation to the multiplicity of plant systems, possible modelling abstractions, and converging models at the theoretical level. The PP approach, as implemented within the ESP-r system (Version 9 Series), is described in terms of its features, applications, limitations and planned developments.
'HVAC component specification booklets,'' Energy conservation in buildings & community systems programme; Final report Annex X, International Energy Agency Operating agent: University of Lie `geA testing and diagnostic procedure for building energy simulation programs
- J Lebrun
- R D Judkoff
IEA 1988. ''HVAC component specification booklets,'' Energy conservation in buildings & community systems programme; Final report Annex X, International Energy Agency. Operating agent: University of Lie `ge, J. Lebrun Judkoff, R.D. 1994. ''A testing and diagnostic procedure for building energy simulation programs,'' in Proc. BEP '94 Conference Facing the Future , York, April 6-8, Building Environmental Performance Analysis Club (BEPAC), York (UK)
Posted by: adjustabletemperature.blogspot.com
Source: https://www.researchgate.net/publication/2593981_Current_Building_Systems_Modelling_Potential_of_ESP-r
