Tools
Three calculators for hot-gas defrost in ammonia refrigeration, based on the correlations of the IIAR design-basis studies. They estimate the hot-gas flow below which no shock forms, the shock to be expected at a given flow, and how slowly a motorized valve should open.
Enter the pipe size, the evaporating temperature and the hot-gas flow. The answer to the first question decides which study applies, as explained on the Guidance page. Each result lists the assumptions behind it.
1. Critical hot-gas flow
The flow below which the gas cannot push a pre-formed slug along: the slug collapses under its own weight first. In the simulations, no shock formed at half this flow and a shock formed in every case at twice this flow.
The flow below which the gas cannot raise the liquid layer into a slug that fills the pipe.
| US units | SI units |
|---|
Assumptions
- Saturated ammonia; gas and liquid densities at the evaporating temperature.
- Densimetric Froude number ρgUg²/(ρlgD) at the critical flow: 0.01.
- The true critical value lies between half and twice the calculated flow.
- Saturated liquid at the evaporating temperature; gas at the evaporating pressure and the hot-gas inlet temperature.
- Modified Taitel–Dukler criterion of Martin et al. (2007), with the liquid filling half the pipe diameter.
Source: IIAR Technical Paper 2025 RR1, Eq. (3).
Source: Martin et al. (2007), ASHRAE 970-RP; IIAR research report (2022).
2. Shock estimate
The slug Joukowsky number Ju = ρl cl vslug is the pressure rise when the slug, at its average speed, is stopped at a closed end. It is the measure of the shock potential used in the studies (see Research).
| US units | SI units |
|---|
Assumptions
- Saturated ammonia, properties at the evaporating temperature.
- A pre-formed slug about ten diameters long sits in a trap and is driven by the hot gas into a horizontal pipe about 100 diameters long, closed at the end, with a thin liquid layer (0.2 D).
- Fitted on pipe diameters of 1 to 10 in. and evaporating temperatures of –45 to –20 °F; checked on 12 and 16 in. (average error 14 %).
- Near the critical flow, the result is reduced smoothly to zero (tanh threshold, width 0.2 ṁcrit).
- Saturated ammonia, properties at the evaporating temperature.
- A layer of liquid filling half the pipe diameter, the case that gave the strongest shocks; used here as the conservative assumption.
- Hot gas enters through an elbow into a horizontal pipe closed by an end cap; fast-acting valve.
- Fitted on pipe diameters of 2 to 6 in., L/D of 27 to 120, evaporating temperatures of –45 to –5 °F and hot-gas temperatures of 18 to 52 °F.
- Under-predicts where a slug is already formed in a trap (Wiencke accident: 1,270 against 3,500 psia).
Source: IIAR research report (2024), final correlation.
Source: IIAR research report (2022), correlation for h/D = 50 %.
3. Opening time of a motorized valve
A motorized valve removes most of the shock if it opens much more slowly than the slug needs to travel the length of the pipework. For safety, the opening time is taken as ten times the travel time.
| US units | SI units |
|---|
Assumptions
- Slug velocity from the shock estimate above: vslug = Ju / (ρl cl), with the correlation for the case chosen (pre-formed slug or liquid layer).
- Opening time = 10 × L / vslug.
- Derived with the average opening characteristic of a common motorized valve type (ICM).
- Tested in simulations of four pipes with a liquid layer and of the Wiencke accident with a pre-formed slug.
Source: IIAR 2022 Technical Paper #1, motorized valves.
The results are for orientation in design and review. The correlations apply only within the assumptions listed with each result.
For a particular system, a CFD simulation of the actual pipework and operating sequence can be carried out. Get in touch.