Liquid Desander – Hydraulic Throughput Model(s) (B-FSM060)
We’ve discussed solids separation performance (D50, D98, separation curve, etc.) for desanders, and now we’ll discuss hydraulic throughput. This is step number two from post B-FSM-054. The size (diameter) of desander and number of units operating in parallel determine the pressure drop for a given flow rate.
We are discussing hydraulic – or liquid only – flow at this time. The addition of free-gas will be discussed in future post when we get to multiphase desanding.
The basic hydraulic throughput model is derived from Plitt (see references), and is shown below.
For liquid-only flow (e.g. gas void fraction = 0%), the flow rate (Q) shows a power law relationship with pressure drop (ΔP). This relationship has two constants that must be determined experimentally for each specific desander geometry – a and b. The power factor (b) can be normalized to 0.5 – thus the relationship is flow rate is proportional to square root of pressure drop.
This relationship can be rearranged and simplified to the form of ΔP equals the square of Q/k. k is termed the k-factor and measured experimentally. Many desander suppliers will publish their k-factor on their capacity curves.
Other process factors will affect throughput (pressure drop). These include volume fraction of solids suspended in the liquid (c) and viscosity of the fluid (µ). These are secondary factors and with suspensions of dilute sand in water show negligible effect. When the carrier fluid is highly viscous (>10 cP) some effect on capacity is shown.
The next article will discuss hydraulic capacity, turndown, and performance of liner style desanders.
References
- Plitt, L.R., “A mathematical model of the hydrocyclone classifier”, CIM Bulletin, December, 1976, pp. 115-123.
