It is important to distinguish between two fundamental concepts: the NPSHr (Net Positive Suction Head required) stated on a pump nameplate **does not change with the density of the fluid**; it is solely related to the pump impeller design, rotational speed and flow rate. In contrast, the NPSHa (Net Positive Suction Head available) for a system is determined by the combined effects of the fluid density, saturated vapour pressure, pipework resistance and static head at the liquid level; this is the starting point for the entire phenomenon.
When the density of the concentrate rises gradually, many people focus solely on density as a single variable, overlooking the concurrent changes in saturated vapour pressure, dissolved gas carrying capacity and boundary layer desorption behaviour within high-salt systems. As these multiple factors combine, NPSHa does not decline at a constant rate but gradually approaches a critical threshold; once a certain critical point is crossed, cavitation occurs suddenly, without any gradual warning.
Let us first consider the changes brought about by the static pressure term. In the NPSHa formula, the static pressure contribution is ρgh. As density ρ increases, the absolute value of the liquid-phase static pressure at the same liquid level increases. This often leads to a habitual assumption in the field that higher density equates to higher static pressure, making cavitation less likely—and this is precisely where the trap lies. However, the saturated vapour pressure of highly concentrated salt solutions does not decrease linearly with density. In low-temperature evaporation systems operating under negative pressure, the boiling point elevation of the salt solution continues to increase; consequently, at the same operating temperature, the saturated vapour pressure of the solution continues to decrease. This creates a trade-off: whilst increased density leads to a gain in static pressure, the rise in boiling point results in a gain from reduced vapour pressure; the two partially offset each other, meaning the actual increase in the system’s safety margin is far smaller than might be imagined. At the same time, the solubility of non-condensable gases, light components and microbubbles in the liquid phase continues to decrease as salt concentration rises. The salt-out effect gradually forces gases originally dissolved in the liquid phase out, transforming them into free microbubbles. These free bubbles act as a latent trigger; whilst the density is rising slowly, they remain dispersed and suspended, having not yet accumulated at the pump inlet.
Throughout the entire period of gradual density increase, the NPSHa fluctuates slightly within a very narrow range, slowly approaching the NPSHr. During this stage, the bubbles are dispersed throughout the flow field; once they enter the impeller, they can be carried away by the fluid and do not form continuous vapour cavities. There are no obvious abnormalities in the pump’s vibration, noise or current, and no clear signs are visible on the instruments; this constitutes the latent stage.
The core factor that actually triggers sudden cavitation is **the coupling of local supersaturation in high-salt media with sudden pressure changes**. The pump suction inlet passage and the leading edge of the impeller inherently contain pressure troughs, where the pressure is lower than that of the main flow in the pipework. When the density of the concentrated liquid reaches a certain critical value and the supersaturation of the salts is already very high, and the liquid phase itself carries a large number of microcrystalline nuclei and desorbed microbubbles; at this point, even the slightest disturbance—be it a minor fluctuation in circulation flow rate, a slight drift in the pressure differential between stages, an instantaneous influx of bubbles caused by local desorption of the gas film from the heat transfer surface, or a slight oscillation in the liquid level—will cause the pressure in the low-pressure zone at the suction inlet to instantly drop below the saturated vapour pressure for that operating condition. This is not a case of density increasing at a constant rate and gradually eroding the safety margin; rather, for a considerable period beforehand, the system remains in a metastable state where it is ‘barely holding on’. The window for this metastable state is extremely narrow, and the disturbance threshold is very low; an almost imperceptible external fluctuation is sufficient to directly disrupt the equilibrium.
There is another characteristic that distinguishes this system entirely from ordinary fresh water: the nucleation threshold for cavitation in a brine system is far lower than in fresh water. For cavitation to occur in fresh water, very low pressure is required to separate the liquid phase and generate cavitation bubbles; however, highly concentrated brine contains a persistent abundance of microcrystals, colloids and organic flocs, all of which serve as ready-made cavitation nuclei, eliminating the need for additional energy to create new gas-liquid interfaces. Once the pressure conditions are met, cavitation bubbles erupt instantaneously in large numbers, rather than growing gradually. Cavitation in pure water typically involves a gradual process, beginning with the appearance of a small number of bubbles and a slow increase in noise; in high-salinity concentrated solutions, this transitional phase is absent, resulting in a direct transition from near-zero cavitation to large-scale cavitation, which manifests as a sudden, unannounced event.
There are also additional factors that are easily overlooked. As the concentration ratio increases, the viscosity of the liquid phase rises in tandem, causing the frictional resistance along the suction pipework to rise gradually; this resistance continuously depletes the NPSHa. The positive contribution to static pressure provided by density is gradually offset by several negative factors: rising viscosity, increasing pipework resistance, and the continuous release of dissolved gas. The positive gains are progressively eroded, causing NPSHa to shift from a slow, modest rise to a gradual decline; this transition zone is, in itself, very difficult to monitor and detect. By the time NPSHa and NPSHr are almost at the same level, any disturbance in the flow field will trigger cavitation.
There is also a lag effect, which is linked to the retention of non-condensable gases and the microbubble film on heat transfer surfaces, as discussed earlier. Microbubbles on the heat transfer surface are not released uniformly and continuously; rather, they often accumulate to a certain extent before detaching in clusters, which are then intermittently swept into the pump inlet by the circulating fluid. When the density is relatively low, a small number of bubbles entering the pump will not induce continuous cavitation; however, when the density is already approaching critical operating conditions, an influx of bubbles of the same volume will act as a trigger, instantly forming extensive cavitation cavities and causing the pump’s operating conditions to deteriorate precipitously.