CDS-100 – generiskt exempel på modellbaserad systemanalys

CDS-100 visar hur ett kylsystem kan återskapas som en interaktiv modell för att analysera driftfall, fel, reglering, larm och systemdynamik. Exemplet är generiskt, men metodiken är relevant för industriella kylsystem, pumpsystem, värmesystem och andra tekniska processystem.

Problem

Tekniska system kan bete sig oväntat när flöde, värmeöverföring, pumpar, ventiler och reglering samverkar. Vid driftstörningar räcker det ofta inte med statiska ritningar eller enstaka mätvärden.

Metod

Systemet byggs som en fysikbaserad modell med tydliga komponenter, mätpunkter, styrsignaler, scenarier och antaganden. Modellen gör det möjligt att testa hypoteser och jämföra driftfall.

Resultat

Resultatet kan användas för felsökning, utbildning, scenariotestning, teknisk kommunikation och beslutsunderlag inför åtgärd eller investering.

Möjlig kundnytta

Snabbare förståelse av systemets dynamik.
Bättre felsökning vid avvikande drift.
Test av fel och scenarier utan påverkan på verklig anläggning.
Tydligare kommunikation mellan drift, underhåll och ingenjörer.
Pedagogiskt underlag för utbildning och intern kompetenshöjning.
Underlag för vidare modellering på högre fidelitetsnivå.

CDS-100 är ett generiskt demonstrationssystem och inte ett konstruktions- eller dimensioneringsunderlag.

← Back to Antranova

3D view — drag to rotate · scroll to zoom

Flow path: V-101 → P-101/P-102 → CV-101 → HX-101 → return → V-101

P&ID — live process view

Trends

Measurements

TT-101

35.0

°C

FT-101

25.0

m³/h

Cooling Power

200

Sim Time

LAH-101 — T ≥ 45°C

LAHH-101 — T ≥ 55°C

Pumps & Valve

CV-101 opening 80%

HX capacity 100%

Heat load 200 kW

Setpoint 35°C

PID controller On

PID Controller

Kp 0.1000

Ki 0.0300

Kd 0.0000

Bias 0.50

PID output50%

I-term0.000

Quick Scenarios

Saved States

No saved states

Event Log

Process description

CDS-100 is a single-loop cooling demonstration system. Water stored in vessel V-101 (volume V = 5 m³) is continuously heated by a process heat load Qₙ and cooled by a shell-and-tube heat exchanger HX-101.

Energy balance — tank temperature

ρ · V · cₙ · dT/dt = Qₙ − Qₕₕₓₙ

ρ = 1000 kg/m³ — water density
cₙ = 4180 J/(kg·K) — specific heat
V = 5 m³ — tank volume → thermal mass M·cₙ = 20.9 MJ/K
Qₙ = adjustable heat load (50–400 kW)
Qₕₕₓₙ = heat removed by HX-101

Flow

Each running pump delivers nominal flow Fₙ = 25 m³/h through CV-101. Two pumps in parallel double the flow. Flow is zero when no pump runs or CV-101 is fully closed.

F = nₘ · Fₙ · x𝐶𝑉 [kg/s]

Heat exchanger — NTU effectiveness method

HX-101 is modelled using the ε-NTU method for a single-pass heat exchanger with cooling water inlet temperature T𝑢 = 20°C:

UA = UA₀ · h𝑓 · (F/Fₙ)^0.8
ε = 1 − exp(−UA / Cₙ)
Qₕₕₓₙ = ε · Cₙ · (T − T𝑢)

UA₀ = 13 333 W/K — design conductance at nominal flow
h𝑓 = HX capacity factor (1.0 = clean, <1 = fouled)
0.8 exponent — Dittus-Boelter turbulent convection dependence
Cₙ = F · cₙ — process stream heat capacity rate [W/K]

Alarms

LAH-101: T ≥ 45°C — high temperature warning
LAHH-101: T ≥ 55°C — high-high temperature, requires immediate action

Instrumentation

TT-101: Process temperature in V-101. Primary input to TIC-101. Also displayed as a readout tag next to the transmitter circle in the P&ID.
FT-101: Volumetric flow rate in the pump discharge header [m³/h]. Computed from pump count and CV-101 position.
TT-102: Process return temperature after HX-101, computed as T₂ = T − Q_cool/(ṁ·c_p). Represents the temperature entering the top of V-101.

Cooling water side

The cooling water (CW) enters HX-101 at T_u = 20°C. The outlet temperature is estimated assuming a nominal CW circulation rate of 30 m³/h (8.33 kg/s):

T_CW,out = T_u + Q_cool / (ṁ_CW · c_p)

At 200 kW cooling duty: T_CW,out ≈ 20 + 200 000 / (8.33 × 4180) ≈ 25.7°C
CW outlet temperature is displayed in the P&ID next to the CW OUT arrow.

PID controller — TIC-101

TIC-101 controls the tank temperature T by adjusting the position of control valve CV-101. The controller acts in reverse: when T rises above setpoint, the valve opens further to increase cooling flow.

PID equation (positional form, discrete)

u(t) = bias + Kₙ · e(t) + K𝑖 · ∫e dt + K𝑑 · de/dt

e(t) = T(t) − Tⱼₙ (error, positive when too hot)

Parameters

Parameter	Symbol	Default	Effect
Bias	b	0.50	Valve position at zero error — set to expected steady-state opening
Proportional gain	Kₙ	0.10	Larger → faster response, smaller offset; too large → oscillation
Integral gain	K𝑖	0.030	Eliminates steady-state offset; too large → slow oscillation / windup
Derivative gain	K𝑑	0.000	Damps fast changes; amplifies measurement noise — use cautiously

Anti-windup

The simulator uses conditional integration: the integral accumulates only when the output u is within [0, 1]. When the valve is saturated (fully open or fully closed), the integral is frozen unless accumulating in that direction would reduce saturation. This prevents integral windup during pump trips and large setpoint steps.

Tuning guide

Temperature oscillates slowly after a setpoint step → K𝑖 is too large. Reduce K𝑖 by 30–50%.
Temperature never reaches setpoint (persistent offset) → K𝑖 is too small, or bias is far from the steady-state operating point. Increase K𝑖 or adjust bias.
Valve hunts rapidly (high-frequency oscillation) → Kₙ is too large. Reduce Kₙ by 50%.
Response is slow after a pump trip → Increase Kₙ. The large thermal mass (≈21 MJ/K) means the system is naturally slow.
Derivative makes the valve jitter → Set K𝑑 = 0. The process has no significant measurement noise but K𝑑 > 0.01 will still interact poorly with the discrete integration step.